bootstrap with serialized mach

.gitignore

@@ -1,6 +1,7 @@
 .git/
 .obj/
-website/
+website/public/
+website/.hugo_build.lock
 bin/
 build/
 *.zip
@@ -15,6 +16,7 @@ build/
 source/shaders/*.h
 .DS_Store
 *.html
+!website/themes/**/*.html
 .vscode
 *.icns
 icon.ico

CLAUDE.md

@@ -1,25 +1,123 @@
-# Code style
-All code is done with 2 spaces for indentation.
+# ƿit (pit) Language Project
 
-For cell script and its integration files, objects are preferred over classes, and preferrably limited use of prototypes, make objects sendable between actors (.ce files).
+## Building
 
-## cell script format
-Cell script files end in .ce or .cm. Cell script is similar to Javascript but with some differences.
+Recompile after changes: `make`
+Bootstrap from scratch (first time): `make bootstrap`
+Run `cell --help` to see all CLI flags.
 
-Variables are delcared with 'var'. Var behaves like let.
-Constants are declared with 'def'.
-!= and == are strict, there is no !== or ===.
-There is no undefined, only null.
-There are no classes, only objects and prototypes.
-Prefer backticks for string interpolation. Otherwise, convering non strings with the text() function is required.
-Everything should be lowercase.
+## Code Style
 
-There are no arraybuffers, only blobs, which work with bits. They must be stoned like stone(blob) before being read from.
+All code uses 2 spaces for indentation. K&R style for C and Javascript.
 
-## c format
-For cell script integration files, everything should be declared static that can be. Most don't have headers at all. Files in a package are not shared between packages.
+## ƿit Script Quick Reference
 
-There is no undefined, so JS_IsNull and JS_NULL should be used only.
+ƿit script files: `.ce` (actors) and `.cm` (modules). The syntax is similar to JavaScript with important differences listed below.
 
-## how module loading is done in cell script
-Within a package, a c file, if using the correct macros (CELL_USE_FUNCS etc), will be loaded as a module with its name; so png.c inside ac package is loaded as <package>/png, giving you access to its functions.
+### Key Differences from JavaScript
+
+- `var` (mutable) and `def` (constant) — no `let` or `const`
+- `==` and `!=` are strict (no `===` or `!==`)
+- No `undefined` — only `null`
+- No classes — only objects and prototypes (`meme()`, `proto()`, `isa()`)
+- No `for...in`, `for...of`, spread (`...`), rest params, or default params
+- No named function declarations — use `var fn = function() {}` or arrow functions
+- Variables must be declared at function body level only (not in if/while/for/blocks)
+- All variables must be initialized at declaration (`var x` alone is an error; use `var x = null`)
+- No `try`/`catch`/`throw` — use `disrupt`/`disruption`
+- No arraybuffers — only `blob` (works with bits; must `stone(blob)` before reading)
+- Identifiers can contain `?` and `!` (e.g., `nil?`, `set!`, `is?valid`)
+- Prefer backticks for string interpolation; otherwise use `text()` to convert non-strings
+- Everything should be lowercase
+
+### Intrinsic Functions (always available, no `use()` needed)
+
+The creator functions are **polymorphic** — behavior depends on argument types:
+
+- `array(number)` — create array of size N filled with null
+- `array(number, value_or_fn)` — create array with initial values
+- `array(array)` — copy array
+- `array(array, fn)` — map
+- `array(array, array)` — concatenate
+- `array(array, from, to)` — slice
+- `array(record)` — get keys as array of text
+- **`array(text)` — split text into individual characters** (e.g., `array("hello")` → `["h","e","l","l","o"]`)
+- `array(text, separator)` — split by separator
+- `array(text, length)` — split into chunks of length
+
+- `text(array, separator)` — join array into text
+- `text(number)` or `text(number, radix)` — number to text
+- `text(text, from, to)` — substring
+
+- `number(text)` or `number(text, radix)` — parse text to number
+- `number(logical)` — boolean to number
+
+- `record(record)` — copy
+- `record(record, another)` — merge
+- `record(array_of_keys)` — create record from keys
+
+Other key intrinsics: `length()`, `stone()`, `is_stone()`, `print()`, `filter()`, `find()`, `reduce()`, `sort()`, `reverse()`, `some()`, `every()`, `starts_with()`, `ends_with()`, `meme()`, `proto()`, `isa()`, `splat()`, `apply()`, `extract()`, `replace()`, `search()`, `format()`, `lower()`, `upper()`, `trim()`
+
+Sensory functions: `is_array()`, `is_text()`, `is_number()`, `is_object()`, `is_function()`, `is_null()`, `is_logical()`, `is_integer()`, `is_stone()`, etc.
+
+### Standard Library (loaded with `use()`)
+
+- `blob` — binary data (bits, not bytes)
+- `time` — time constants and conversions
+- `math` — trig, logarithms, roots (`math/radians`, `math/turns`)
+- `json` — JSON encoding/decoding
+- `random` — random number generation
+
+### Actor Model
+
+- `.ce` files are actors (independent execution units, don't return values)
+- `.cm` files are modules (return a value, cached and frozen)
+- Actors never share memory; communicate via `$send()` message passing
+- Actor intrinsics start with `$`: `$me`, `$stop()`, `$send()`, `$start()`, `$delay()`, `$receiver()`, `$clock()`, `$portal()`, `$contact()`, `$couple()`, `$unneeded()`, `$connection()`, `$time_limit()`
+
+### Requestors (async composition)
+
+`sequence()`, `parallel()`, `race()`, `fallback()` — compose asynchronous operations. See docs/requestors.md.
+
+### Error Handling
+
+```javascript
+var fn = function() {
+  disrupt  // bare keyword, no value
+} disruption {
+  // handle error; can re-raise with disrupt
+}
+```
+
+### Push/Pop Syntax
+
+```javascript
+var a = [1, 2]
+a[] = 3       // push: [1, 2, 3]
+var v = a[]   // pop: v is 3, a is [1, 2]
+```
+
+## C Integration
+
+- Declare everything `static` that can be
+- Most files don't have headers; files in a package are not shared between packages
+- No undefined in C API: use `JS_IsNull` and `JS_NULL` only
+- A C file with correct macros (`CELL_USE_FUNCS` etc) is loaded as a module by its name (e.g., `png.c` in a package → `use('<package>/png')`)
+
+## Project Layout
+
+- `source/` — C source for the cell runtime and CLI
+- `docs/` — master documentation (Markdown), reflected on the website
+- `website/` — Hugo site; theme at `website/themes/knr/`
+- `internal/` — internal ƿit scripts (engine.cm etc.)
+- `packages/` — core packages
+- `Makefile` — build system (`make` to rebuild, `make bootstrap` for first build)
+
+## Documentation
+
+The `docs/` folder is the single source of truth. The website at `website/` mounts it via Hugo. Key files:
+- `docs/language.md` — language syntax reference
+- `docs/functions.md` — all built-in intrinsic functions
+- `docs/actors.md` — actor model and actor intrinsics
+- `docs/requestors.md` — async requestor pattern
+- `docs/library/*.md` — intrinsic type reference (text, number, array, object) and standard library modules

Makefile

@@ -58,7 +58,7 @@ static:
 # Bootstrap: build cell from scratch using meson (only needed once)
 # Also installs core scripts to ~/.cell/core
 bootstrap: 
-	meson setup build_bootstrap -Dbuildtype=debug -Db_sanitize=address
+	meson setup build_bootstrap -Dbuildtype=debugoptimized
 	meson compile -C build_bootstrap
 	cp build_bootstrap/cell .
 	cp build_bootstrap/libcell_runtime.dylib .

archive/miniz.c

@@ -382,13 +382,13 @@ static const JSCFunctionListEntry js_reader_funcs[] = {
 JSValue js_miniz_use(JSContext *js)
 {
   JS_NewClassID(&js_reader_class_id);
-  JS_NewClass(JS_GetRuntime(js), js_reader_class_id, &js_reader_class);
+  JS_NewClass(js, js_reader_class_id, &js_reader_class);
   JSValue reader_proto = JS_NewObject(js);
   JS_SetPropertyFunctionList(js, reader_proto, js_reader_funcs, sizeof(js_reader_funcs) / sizeof(JSCFunctionListEntry));
   JS_SetClassProto(js, js_reader_class_id, reader_proto);
 
   JS_NewClassID(&js_writer_class_id);
-  JS_NewClass(JS_GetRuntime(js), js_writer_class_id, &js_writer_class);
+  JS_NewClass(js, js_writer_class_id, &js_writer_class);
   JSValue writer_proto = JS_NewObject(js);
   JS_SetPropertyFunctionList(js, writer_proto, js_writer_funcs, sizeof(js_writer_funcs) / sizeof(JSCFunctionListEntry));
   JS_SetClassProto(js, js_writer_class_id, writer_proto);

bootstrap.ce

@@ -0,0 +1,41 @@
+// bootstrap.ce — regenerate .mach bytecode files consumed by the mach engine
+// usage: cell bootstrap.ce
+
+var fd = use("fd")
+var json = use("json")
+var tokenize = use("tokenize")
+var parse = use("parse")
+var fold = use("fold")
+
+var files = [
+  {src: "tokenize.cm", name: "tokenize", out: "tokenize.mach"},
+  {src: "parse.cm", name: "parse", out: "parse.mach"},
+  {src: "fold.cm", name: "fold", out: "fold.mach"},
+  {src: "mcode.cm", name: "mcode", out: "mcode.mach"},
+  {src: "internal/bootstrap.cm", name: "bootstrap", out: "internal/bootstrap.mach"}
+]
+
+var i = 0
+var entry = null
+var src = null
+var tok_result = null
+var ast = null
+var folded = null
+var ast_json = null
+var bytecode = null
+var f = null
+
+while (i < length(files)) {
+  entry = files[i]
+  src = text(fd.slurp(entry.src))
+  tok_result = tokenize(src, entry.src)
+  ast = parse(tok_result.tokens, src, entry.src, tokenize)
+  folded = fold(ast)
+  ast_json = json.encode(folded)
+  bytecode = mach_compile_ast(entry.name, ast_json)
+  f = fd.open(entry.out, "w")
+  fd.write(f, bytecode)
+  fd.close(f)
+  print(`wrote ${entry.out}`)
+  i = i + 1
+}

debug/js.c

@@ -13,7 +13,6 @@ JSC_CCALL(os_calc_mem,
   JS_SetPropertyStr(js,ret,"memory_used_size",number2js(js,mu.memory_used_size));
   JS_SetPropertyStr(js,ret,"malloc_count",number2js(js,mu.malloc_count));
   JS_SetPropertyStr(js,ret,"memory_used_count",number2js(js,mu.memory_used_count));
-  /* atom_count and atom_size removed - atoms are now just strings */
   JS_SetPropertyStr(js,ret,"str_count",number2js(js,mu.str_count));
   JS_SetPropertyStr(js,ret,"str_size",number2js(js,mu.str_size));
   JS_SetPropertyStr(js,ret,"obj_count",number2js(js,mu.obj_count));
@@ -35,49 +34,6 @@ JSC_CCALL(os_calc_mem,
   JS_SetPropertyStr(js,ret,"binary_object_size",number2js(js,mu.binary_object_size));
 )
 
-// Evaluate a string of JavaScript code in the current QuickJS context.
-JSC_SSCALL(os_eval,
-  if (!str2) return JS_ThrowReferenceError(js, "Second argument should be the script.");
-  if (!str) return JS_ThrowReferenceError(js, "First argument should be the name of the script.");
-  JSValue bytecode = JS_Compile(js, str2, strlen(str2), str);
-  if (JS_IsException(bytecode)) return bytecode;
-  ret = JS_Integrate(js, bytecode, JS_NULL);
-)
-
-// Compile a string of JavaScript code into a function object.
-JSC_SSCALL(js_compile,
-  if (!str2) return JS_ThrowReferenceError(js, "Second argument should be the script.");
-  if (!str) return JS_ThrowReferenceError(js, "First argument should be the name of the script.");
-  ret = JS_Compile(js, str2, strlen(str2), str);
-)
-
-// Link compiled bytecode with environment and execute.
-JSC_CCALL(js_integrate,
-  JSValue env = (argc > 1 && !JS_IsNull(argv[1])) ? argv[1] : JS_NULL;
-  ret = JS_Integrate(js, argv[0], env);
-)
-
-// Compile a function object into a bytecode blob.
-JSC_CCALL(js_compile_blob,
-  size_t size;
-  uint8_t *data = JS_WriteObject(js, &size, argv[0], JS_WRITE_OBJ_BYTECODE);
-  if (!data) {
-    return JS_ThrowInternalError(js, "Failed to serialize bytecode");
-  }
-  ret = js_new_blob_stoned_copy(js, data, size);
-  js_free(js, data);
-)
-
-// Compile a bytecode blob into a function object.
-JSC_CCALL(js_compile_unblob,
-  size_t size;
-  void *data = js_get_blob_data(js, &size, argv[0]);
-  if (data == -1) return JS_EXCEPTION;
-  if (!data) return JS_ThrowReferenceError(js, "No data present in blob.");
-  
-  return JS_ReadObject(js, data, size, JS_READ_OBJ_BYTECODE);
-)
-
 // Disassemble a function object into a string.
 JSC_CCALL(js_disassemble,
   return js_debugger_fn_bytecode(js, argv[0]);
@@ -92,11 +48,6 @@ static const JSCFunctionListEntry js_js_funcs[] = {
   MIST_FUNC_DEF(os, calc_mem, 0),
   MIST_FUNC_DEF(os, mem_limit, 1),
   MIST_FUNC_DEF(os, max_stacksize, 1),
-  MIST_FUNC_DEF(os, eval, 2),
-  MIST_FUNC_DEF(js, compile, 2),
-  MIST_FUNC_DEF(js, integrate, 2),
-  MIST_FUNC_DEF(js, compile_blob, 1),
-  MIST_FUNC_DEF(js, compile_unblob, 1),
   MIST_FUNC_DEF(js, disassemble, 1),
   MIST_FUNC_DEF(js, fn_info, 1),
 };
@@ -105,4 +56,4 @@ JSValue js_js_use(JSContext *js) {
   JSValue mod = JS_NewObject(js);
   JS_SetPropertyFunctionList(js,mod,js_js_funcs,countof(js_js_funcs));
   return mod;
-}
+}

docs/.pages

@@ -1,9 +0,0 @@
-nav:
-  - index.md
-  - cellscript.md
-  - actors.md
-  - packages.md
-  - cli.md
-  - c-modules.md
-  - Standard Library: library
-

docs/_index.md

@@ -0,0 +1,90 @@
+---
+title: "Documentation"
+description: "ƿit language documentation"
+type: "docs"
+---
+
+![image](/images/wizard.png)
+
+ƿit is an actor-based scripting language for building concurrent applications. It combines a familiar C-like syntax with the actor model of computation, optimized for low memory usage and simplicity.
+
+## Key Features
+
+- **Actor Model** — isolated memory, message passing, no shared state
+- **Immutability** — `stone()` makes values permanently frozen
+- **Prototype Inheritance** — objects without classes
+- **C Integration** — seamlessly extend with native code
+- **Cross-Platform** — deploy to desktop, web, and embedded
+
+## Quick Start
+
+```javascript
+// hello.ce - A simple actor
+print("Hello, ƿit!")
+$stop()
+```
+
+```bash
+pit hello
+```
+
+## Language
+
+- [**ƿit Language**](/docs/language/) — syntax, types, and operators
+- [**Actors and Modules**](/docs/actors/) — the execution model
+- [**Requestors**](/docs/requestors/) — asynchronous composition
+- [**Packages**](/docs/packages/) — code organization and sharing
+
+## Reference
+
+- [**Built-in Functions**](/docs/functions/) — intrinsics reference
+- [text](/docs/library/text/) — text conversion and manipulation
+- [number](/docs/library/number/) — numeric conversion and operations
+- [array](/docs/library/array/) — array creation and manipulation
+- [object](/docs/library/object/) — object creation, prototypes, and serialization
+
+## Standard Library
+
+Modules loaded with `use()`:
+
+- [blob](/docs/library/blob/) — binary data
+- [time](/docs/library/time/) — time and dates
+- [math](/docs/library/math/) — trigonometry and math
+- [json](/docs/library/json/) — JSON encoding/decoding
+- [random](/docs/library/random/) — random numbers
+
+## Tools
+
+- [**Command Line**](/docs/cli/) — the `pit` tool
+- [**Writing C Modules**](/docs/c-modules/) — native extensions
+
+## Architecture
+
+ƿit programs are organized into **packages**. Each package contains:
+
+- **Modules** (`.cm`) — return a value, cached and frozen
+- **Actors** (`.ce`) — run independently, communicate via messages
+- **C files** (`.c`) — compiled to native libraries
+
+Actors never share memory. They communicate by sending messages, which are automatically serialized. This makes concurrent programming safe and predictable.
+
+## Installation
+
+```bash
+# Clone and bootstrap
+git clone https://gitea.pockle.world/john/cell
+cd cell
+make bootstrap
+```
+
+The ƿit shop is stored at `~/.pit/`.
+
+## Development
+
+After making changes, recompile with:
+
+```bash
+make
+```
+
+Run `cell --help` to see all available CLI flags.

docs/actors.md

@@ -1,10 +1,15 @@
-# Actors and Modules
+---
+title: "Actors and Modules"
+description: "The ƿit execution model"
+weight: 20
+type: "docs"
+---
 
-Cell organizes code into two types of scripts: **modules** (`.cm`) and **actors** (`.ce`).
+ƿit organizes code into two types of scripts: **modules** (`.cm`) and **actors** (`.ce`).
 
 ## The Actor Model
 
-Cell is built on the actor model of computation. Each actor:
+ƿit is built on the actor model of computation. Each actor:
 
 - Has its own **isolated memory** — actors never share state
 - Runs to completion each **turn** — no preemption
@@ -21,13 +26,13 @@ A module is a script that **returns a value**. The returned value is cached and
 // math_utils.cm
 var math = use('math/radians')
 
-function distance(x1, y1, x2, y2) {
+var distance = function(x1, y1, x2, y2) {
   var dx = x2 - x1
   var dy = y2 - y1
   return math.sqrt(dx * dx + dy * dy)
 }
 
-function midpoint(x1, y1, x2, y2) {
+var midpoint = function(x1, y1, x2, y2) {
   return {
     x: (x1 + x2) / 2,
     y: (y1 + y2) / 2
@@ -60,12 +65,12 @@ An actor is a script that **does not return a value**. It runs as an independent
 
 ```javascript
 // worker.ce
-log.console("Worker started")
+print("Worker started")
 
-$on_message = function(msg) {
-  log.console("Received:", msg)
+$receiver(function(msg, reply) {
+  print("Received:", msg)
   // Process message...
-}
+})
 ```
 
 **Key properties:**
@@ -83,7 +88,7 @@ Actors have access to special functions prefixed with `$`:
 Reference to the current actor.
 
 ```javascript
-log.console($me)  // actor reference
+print($me)  // actor reference
 ```
 
 ### $stop()
@@ -100,7 +105,7 @@ Send a message to another actor.
 
 ```javascript
 $send(other_actor, {type: "ping", data: 42}, function(reply) {
-  log.console("Got reply:", reply)
+  print("Got reply:", reply)
 })
 ```
 
@@ -112,7 +117,7 @@ Start a new actor from a script.
 
 ```javascript
 $start(function(new_actor) {
-  log.console("Started:", new_actor)
+  print("Started:", new_actor)
 }, "worker")
 ```
 
@@ -122,7 +127,7 @@ Schedule a callback after a delay.
 
 ```javascript
 $delay(function() {
-  log.console("5 seconds later")
+  print("5 seconds later")
 }, 5)
 ```
 
@@ -169,19 +174,47 @@ $contact(function(connection) {
 
 ### $time_limit(requestor, seconds)
 
-Wrap a requestor with a timeout.
+Wrap a requestor with a timeout. See [Requestors](/docs/requestors/) for details.
 
 ```javascript
 $time_limit(my_requestor, 10)  // 10 second timeout
 ```
 
+### $couple(actor)
+
+Couple the current actor to another actor. When the coupled actor dies, the current actor also dies. Coupling is automatic between an actor and its overling (parent).
+
+```javascript
+$couple(other_actor)
+```
+
+### $unneeded(callback, seconds)
+
+Schedule the actor for removal after a specified time.
+
+```javascript
+$unneeded(function() {
+  // cleanup before removal
+}, 30)
+```
+
+### $connection(callback, actor, config)
+
+Get information about the connection to another actor, such as latency, bandwidth, and activity.
+
+```javascript
+$connection(function(info) {
+  print(info.latency)
+}, other_actor, {})
+```
+
 ## Module Resolution
 
-When you call `use('name')`, Cell searches:
+When you call `use('name')`, ƿit searches:
 
 1. **Current package** — files relative to package root
-2. **Dependencies** — packages declared in `cell.toml`
-3. **Core** — built-in Cell modules
+2. **Dependencies** — packages declared in `pit.toml`
+3. **Core** — built-in ƿit modules
 
 ```javascript
 // From within package 'myapp':
@@ -199,14 +232,14 @@ Files starting with underscore (`_helper.cm`) are private to the package.
 // main.ce - Entry point
 var config = use('config')
 
-log.console("Starting application...")
+print("Starting application...")
 
 $start(function(worker) {
   $send(worker, {task: "process", data: [1, 2, 3]})
 }, "worker")
 
 $delay(function() {
-  log.console("Shutting down")
+  print("Shutting down")
   $stop()
 }, 10)
 ```

docs/c-modules.md

@@ -1,6 +1,11 @@
-# Writing C Modules
+---
+title: "Writing C Modules"
+description: "Extending ƿit with native code"
+weight: 50
+type: "docs"
+---
 
-Cell makes it easy to extend functionality with C code. C files in a package are compiled into a dynamic library and can be imported like any other module.
+ƿit makes it easy to extend functionality with C code. C files in a package are compiled into a dynamic library and can be imported like any other module.
 
 ## Basic Structure
 
@@ -45,12 +50,12 @@ Where:
 - `<filename>` is the C file name without extension
 
 Examples:
-- `mypackage/math.c` → `js_mypackage_math_use`
-- `gitea.pockle.world/john/lib/render.c` → `js_gitea_pockle_world_john_lib_render_use`
+- `mypackage/math.c` -> `js_mypackage_math_use`
+- `gitea.pockle.world/john/lib/render.c` -> `js_gitea_pockle_world_john_lib_render_use`
 
 ## Required Headers
 
-Include `cell.h` for all Cell integration:
+Include `cell.h` for all ƿit integration:
 
 ```c
 #include "cell.h"
@@ -63,7 +68,7 @@ This provides:
 
 ## Conversion Functions
 
-### JavaScript ↔ C
+### JavaScript <-> C
 
 ```c
 // Numbers
@@ -201,7 +206,7 @@ static const JSCFunctionListEntry js_funcs[] = {
 CELL_USE_FUNCS(js_funcs)
 ```
 
-Usage in Cell:
+Usage in ƿit:
 
 ```javascript
 var vector = use('vector')
@@ -211,7 +216,7 @@ var n = vector.normalize(3, 4) // {x: 0.6, y: 0.8}
 var d = vector.dot(1, 0, 0, 1) // 0
 ```
 
-## Combining C and Cell
+## Combining C and ƿit
 
 A common pattern is to have a C file provide low-level functions and a `.cm` file provide a higher-level API:
 
@@ -224,7 +229,7 @@ A common pattern is to have a C file provide low-level functions and a `.cm` fil
 // vector.cm
 var native = this  // C module passed as 'this'
 
-function Vector(x, y) {
+var Vector = function(x, y) {
   return {x: x, y: y}
 }
 
@@ -244,11 +249,11 @@ return Vector
 C files are automatically compiled when you run:
 
 ```bash
-cell build
-cell update
+pit build
+pit update
 ```
 
-The resulting dynamic library is placed in `~/.cell/lib/`.
+The resulting dynamic library is placed in `~/.pit/lib/`.
 
 ## Platform-Specific Code
 
@@ -260,7 +265,7 @@ audio_playdate.c  # Playdate
 audio_emscripten.c # Web/Emscripten
 ```
 
-Cell selects the appropriate file based on the target platform.
+ƿit selects the appropriate file based on the target platform.
 
 ## Static Declarations
 

docs/cellscript.md

@@ -1,288 +0,0 @@
-# Cell Language
-
-Cell is a scripting language for actor-based programming. It combines a familiar syntax with a prototype-based object system and strict immutability semantics.
-
-## Basics
-
-### Variables and Constants
-
-```javascript
-var x = 10        // mutable variable (block-scoped like let)
-def PI = 3.14159  // constant (cannot be reassigned)
-```
-
-### Data Types
-
-Cell has six fundamental types:
-
-- **number** — DEC64 decimal floating point (no rounding errors)
-- **text** — Unicode strings
-- **logical** — `true` or `false`
-- **null** — the absence of a value (no `undefined`)
-- **array** — ordered, numerically-indexed sequences
-- **object** — key-value records with prototype inheritance
-- **blob** — binary data (bits, not bytes)
-- **function** — first-class callable values
-
-### Literals
-
-```javascript
-// Numbers
-42
-3.14
-1_000_000        // underscores for readability
-
-// Text
-"hello"
-'world'
-`template ${x}`  // string interpolation
-
-// Logical
-true
-false
-
-// Null
-null
-
-// Arrays
-[1, 2, 3]
-["a", "b", "c"]
-
-// Objects
-{name: "cell", version: 1}
-{x: 10, y: 20}
-```
-
-### Operators
-
-```javascript
-// Arithmetic
-+ - * / %
-**              // exponentiation
-
-// Comparison (always strict)
-==              // equals (like === in JS)
-!=              // not equals (like !== in JS)
-< > <= >=
-
-// Logical
-&& || !
-
-// Assignment
-= += -= *= /=
-```
-
-### Control Flow
-
-```javascript
-// Conditionals
-if (x > 0) {
-  log.console("positive")
-} else if (x < 0) {
-  log.console("negative")
-} else {
-  log.console("zero")
-}
-
-// Ternary
-var sign = x > 0 ? 1 : -1
-
-// Loops
-for (var i = 0; i < 10; i++) {
-  log.console(i)
-}
-
-for (var item of items) {
-  log.console(item)
-}
-
-for (var key in obj) {
-  log.console(key, obj[key])
-}
-
-while (condition) {
-  // body
-}
-
-// Control
-break
-continue
-return value
-throw "error message"
-```
-
-### Functions
-
-```javascript
-// Named function
-function add(a, b) {
-  return a + b
-}
-
-// Anonymous function
-var multiply = function(a, b) {
-  return a * b
-}
-
-// Arrow function
-var square = x => x * x
-var sum = (a, b) => a + b
-
-// Rest parameters
-function log_all(...args) {
-  for (var arg of args) log.console(arg)
-}
-
-// Default parameters
-function greet(name, greeting = "Hello") {
-  return `${greeting}, ${name}!`
-}
-```
-
-All closures capture `this` (like arrow functions in JavaScript).
-
-## Arrays
-
-Arrays are **distinct from objects**. They are ordered, numerically-indexed sequences. You cannot add arbitrary string keys to an array.
-
-```javascript
-var arr = [1, 2, 3]
-arr[0]              // 1
-arr[2] = 10         // [1, 2, 10]
-length(arr)         // 3
-
-// Array spread
-var more = [...arr, 4, 5]  // [1, 2, 10, 4, 5]
-```
-
-## Objects
-
-Objects are key-value records with prototype-based inheritance.
-
-```javascript
-var point = {x: 10, y: 20}
-point.x             // 10
-point["y"]          // 20
-
-// Object spread
-var point3d = {...point, z: 30}
-
-// Prototype inheritance
-var colored_point = {__proto__: point, color: "red"}
-colored_point.x     // 10 (inherited)
-```
-
-### Prototypes
-
-```javascript
-// Create object with prototype
-var child = meme(parent)
-
-// Get prototype
-var p = proto(child)
-
-// Check prototype chain
-isa(child, parent)  // true
-```
-
-## Immutability with Stone
-
-The `stone()` function makes values permanently immutable.
-
-```javascript
-var config = stone({
-  debug: true,
-  maxRetries: 3
-})
-
-config.debug = false  // Error! Stone objects cannot be modified
-```
-
-Stone is **deep** — all nested objects and arrays are also frozen. This cannot be reversed.
-
-```javascript
-stone.p(value)  // returns true if value is stone
-```
-
-## Built-in Functions
-
-### length(value)
-
-Returns the length of arrays (elements), text (codepoints), blobs (bits), or functions (arity).
-
-```javascript
-length([1, 2, 3])      // 3
-length("hello")        // 5
-length(function(a,b){}) // 2
-```
-
-### use(path)
-
-Import a module. Returns the cached, stone value.
-
-```javascript
-var math = use('math/radians')
-var json = use('json')
-```
-
-### isa(value, type)
-
-Check type or prototype chain.
-
-```javascript
-is_number(42)        // true
-is_text("hi")        // true
-is_array([1,2])      // true
-is_object({})        // true
-isa(child, parent)     // true if parent is in prototype chain
-```
-
-### reverse(array)
-
-Returns a new array with elements in reverse order.
-
-```javascript
-reverse([1, 2, 3])  // [3, 2, 1]
-```
-
-### logical(value)
-
-Convert to boolean.
-
-```javascript
-logical(0)        // false
-logical(1)        // true
-logical("true")   // true
-logical("false")  // false
-logical(null)     // false
-```
-
-## Logging
-
-```javascript
-log.console("message")   // standard output
-log.error("problem")     // error output
-```
-
-## Pattern Matching
-
-Cell supports regex patterns in string functions, but not standalone regex objects.
-
-```javascript
-text.search("hello world", /world/)
-replace("hello", /l/g, "L")
-```
-
-## Error Handling
-
-```javascript
-try {
-  riskyOperation()
-} catch (e) {
-  log.error(e)
-}
-
-throw "something went wrong"
-```
-
-If an actor has an uncaught error, it crashes.

docs/cli.md

@@ -1,138 +1,143 @@
-# Command Line Interface
+---
+title: "Command Line Interface"
+description: "The pit tool"
+weight: 40
+type: "docs"
+---
 
-Cell provides a command-line interface for managing packages, running scripts, and building applications.
+ƿit provides a command-line interface for managing packages, running scripts, and building applications.
 
 ## Basic Usage
 
 ```bash
-cell <command> [arguments]
+pit <command> [arguments]
 ```
 
 ## Commands
 
-### cell version
+### pit version
 
-Display the Cell version.
+Display the ƿit version.
 
 ```bash
-cell version
+pit version
 # 0.1.0
 ```
 
-### cell install
+### pit install
 
 Install a package to the shop.
 
 ```bash
-cell install gitea.pockle.world/john/prosperon
-cell install /Users/john/local/mypackage  # local path
+pit install gitea.pockle.world/john/prosperon
+pit install /Users/john/local/mypackage  # local path
 ```
 
-### cell update
+### pit update
 
 Update packages from remote sources.
 
 ```bash
-cell update              # update all packages
-cell update <package>    # update specific package
+pit update              # update all packages
+pit update <package>    # update specific package
 ```
 
-### cell remove
+### pit remove
 
 Remove a package from the shop.
 
 ```bash
-cell remove gitea.pockle.world/john/oldpackage
+pit remove gitea.pockle.world/john/oldpackage
 ```
 
-### cell list
+### pit list
 
 List installed packages.
 
 ```bash
-cell list              # list all installed packages
-cell list <package>    # list dependencies of a package
+pit list              # list all installed packages
+pit list <package>    # list dependencies of a package
 ```
 
-### cell ls
+### pit ls
 
 List modules and actors in a package.
 
 ```bash
-cell ls              # list files in current project
-cell ls <package>    # list files in specified package
+pit ls              # list files in current project
+pit ls <package>    # list files in specified package
 ```
 
-### cell build
+### pit build
 
 Build the current package.
 
 ```bash
-cell build
+pit build
 ```
 
-### cell test
+### pit test
 
 Run tests.
 
 ```bash
-cell test            # run tests in current package
-cell test all        # run all tests
-cell test <package>  # run tests in specific package
+pit test            # run tests in current package
+pit test all        # run all tests
+pit test <package>  # run tests in specific package
 ```
 
-### cell link
+### pit link
 
 Manage local package links for development.
 
 ```bash
-cell link add <canonical> <local_path>  # link a package
-cell link list                          # show all links
-cell link delete <canonical>            # remove a link
-cell link clear                         # remove all links
+pit link add <canonical> <local_path>  # link a package
+pit link list                          # show all links
+pit link delete <canonical>            # remove a link
+pit link clear                         # remove all links
 ```
 
-### cell fetch
+### pit fetch
 
 Fetch package sources without extracting.
 
 ```bash
-cell fetch <package>
+pit fetch <package>
 ```
 
-### cell upgrade
+### pit upgrade
 
-Upgrade the Cell installation itself.
+Upgrade the ƿit installation itself.
 
 ```bash
-cell upgrade
+pit upgrade
 ```
 
-### cell clean
+### pit clean
 
 Clean build artifacts.
 
 ```bash
-cell clean
+pit clean
 ```
 
-### cell help
+### pit help
 
 Display help information.
 
 ```bash
-cell help
-cell help <command>
+pit help
+pit help <command>
 ```
 
 ## Running Scripts
 
-Any `.ce` file in the Cell core can be run as a command:
+Any `.ce` file in the ƿit core can be run as a command:
 
 ```bash
-cell version    # runs version.ce
-cell build      # runs build.ce
-cell test       # runs test.ce
+pit version    # runs version.ce
+pit build      # runs build.ce
+pit test       # runs test.ce
 ```
 
 ## Package Locators
@@ -143,16 +148,16 @@ Packages are identified by locators:
 - **Local**: `/absolute/path/to/package`
 
 ```bash
-cell install gitea.pockle.world/john/prosperon
-cell install /Users/john/work/mylib
+pit install gitea.pockle.world/john/prosperon
+pit install /Users/john/work/mylib
 ```
 
 ## Configuration
 
-Cell stores its data in `~/.cell/`:
+ƿit stores its data in `~/.pit/`:
 
 ```
-~/.cell/
+~/.pit/
 ├── packages/       # installed packages
 ├── lib/            # compiled dynamic libraries
 ├── build/          # build cache
@@ -163,7 +168,7 @@ Cell stores its data in `~/.cell/`:
 
 ## Environment
 
-Cell reads the `HOME` environment variable to locate the shop directory.
+ƿit reads the `HOME` environment variable to locate the shop directory.
 
 ## Exit Codes
 

docs/index.md

@@ -1,66 +0,0 @@
-# Cell
-
-![image](wizard.png)
-
-Cell is an actor-based scripting language for building concurrent applications. It combines a familiar C-like syntax with the actor model of computation, optimized for low memory usage and simplicity.
-
-## Key Features
-
-- **Actor Model** — isolated memory, message passing, no shared state
-- **Immutability** — `stone()` makes values permanently frozen
-- **Prototype Inheritance** — objects without classes
-- **C Integration** — seamlessly extend with native code
-- **Cross-Platform** — deploy to desktop, web, and embedded
-
-## Quick Start
-
-```javascript
-// hello.ce - A simple actor
-log.console("Hello, Cell!")
-$stop()
-```
-
-```bash
-cell hello
-```
-
-## Documentation
-
-- [**Cell Language**](cellscript.md) — syntax, types, and built-in functions
-- [**Actors and Modules**](actors.md) — the execution model
-- [**Packages**](packages.md) — code organization and sharing
-- [**Command Line**](cli.md) — the `cell` tool
-- [**Writing C Modules**](c-modules.md) — native extensions
-
-## Standard Library
-
-- [text](library/text.md) — string manipulation
-- [number](library/number.md) — numeric operations (functions are global: `floor()`, `max()`, etc.)
-- [array](library/array.md) — array utilities
-- [object](library/object.md) — object utilities
-- [blob](library/blob.md) — binary data
-- [time](library/time.md) — time and dates
-- [math](library/math.md) — trigonometry and math
-- [json](library/json.md) — JSON encoding/decoding
-- [random](library/random.md) — random numbers
-
-## Architecture
-
-Cell programs are organized into **packages**. Each package contains:
-
-- **Modules** (`.cm`) — return a value, cached and frozen
-- **Actors** (`.ce`) — run independently, communicate via messages
-- **C files** (`.c`) — compiled to native libraries
-
-Actors never share memory. They communicate by sending messages, which are automatically serialized. This makes concurrent programming safe and predictable.
-
-## Installation
-
-```bash
-# Clone and bootstrap
-git clone https://gitea.pockle.world/john/cell
-cd cell
-make bootstrap
-```
-
-The Cell shop is stored at `~/.cell/`.

docs/kim.md

@@ -0,0 +1,94 @@
+---
+title: "Kim Encoding"
+description: "Compact character and count encoding"
+weight: 80
+type: "docs"
+---
+
+Kim is a character and count encoding designed by Douglas Crockford. It encodes Unicode characters and variable-length integers using continuation bytes. Kim is simpler and more compact than UTF-8 for most text.
+
+## Continuation Bytes
+
+The fundamental idea in Kim is the continuation byte:
+
+```
+C  D  D  D  D  D  D  D
+```
+
+- **C** — continue bit. If 1, read another byte. If 0, this is the last byte.
+- **D** (7 bits) — data bits.
+
+To decode: shift the accumulator left by 7 bits, add the 7 data bits. If the continue bit is 1, repeat with the next byte. If 0, the value is complete.
+
+To encode: take the value, emit 7 bits at a time from most significant to least significant, setting the continue bit on all bytes except the last.
+
+## Character Encoding
+
+Kim encodes Unicode codepoints directly as continuation byte sequences:
+
+| Range | Bytes | Characters |
+|-------|-------|------------|
+| U+0000 to U+007F | 1 | ASCII |
+| U+0080 to U+3FFF | 2 | First quarter of BMP |
+| U+4000 to U+10FFFF | 3 | All other Unicode |
+
+Unlike UTF-8, there is no need for surrogate pairs or escapement. Every Unicode character, including emoji and characters from extended planes, is encoded in at most 3 bytes.
+
+### Examples
+
+```
+'A'       (U+0041)  →  41
+'é'       (U+00E9)  →  81 69
+'💩'      (U+1F4A9)  →  87 E9 29
+```
+
+## Count Encoding
+
+Kim is also used for encoding counts (lengths, sizes). The same continuation byte format represents non-negative integers of arbitrary size:
+
+| Range | Bytes |
+|-------|-------|
+| 0 to 127 | 1 |
+| 128 to 16383 | 2 |
+| 16384 to 2097151 | 3 |
+
+## Comparison with UTF-8
+
+| Property | Kim | UTF-8 |
+|----------|-----|-------|
+| ASCII | 1 byte | 1 byte |
+| BMP (first quarter) | 2 bytes | 2-3 bytes |
+| Full Unicode | 3 bytes | 3-4 bytes |
+| Self-synchronizing | No | Yes |
+| Sortable | No | Yes |
+| Simpler to implement | Yes | No |
+| Byte count for counts | Variable (7 bits/byte) | Not applicable |
+
+Kim trades self-synchronization (the ability to find character boundaries from any position) for simplicity and compactness. In practice, Kim text is accessed sequentially, so self-synchronization is not needed.
+
+## Usage in ƿit
+
+Kim is used internally by blobs and by the Nota message format.
+
+### In Blobs
+
+The `blob.write_text` and `blob.read_text` functions use Kim to encode text into binary data:
+
+```javascript
+var blob = use('blob')
+var b = blob.make()
+blob.write_text(b, "hello")  // Kim-encoded length + characters
+stone(b)
+var text = blob.read_text(b, 0)  // "hello"
+```
+
+### In Nota
+
+Nota uses Kim for two purposes:
+
+1. **Counts** — array lengths, text lengths, blob sizes, record pair counts
+2. **Characters** — text content within Nota messages
+
+The preamble byte of each Nota value incorporates the first few bits of a Kim-encoded count, with the continue bit indicating whether more bytes follow.
+
+See [Nota Format](#nota) for the full specification.

docs/language.md

@@ -0,0 +1,649 @@
+---
+title: "ƿit Language"
+description: "Syntax, types, operators, and built-in functions"
+weight: 10
+type: "docs"
+---
+
+ƿit is a scripting language for actor-based programming. It combines a familiar syntax with a prototype-based object system and strict immutability semantics.
+
+## Basics
+
+### Variables and Constants
+
+Variables are declared with `var`, constants with `def`. All declarations must be initialized and must appear at the function body level — not inside `if`, `while`, `for`, or bare `{}` blocks.
+
+```javascript
+var x = 10
+var name = "pit"
+var empty = null
+
+def PI = 3.14159   // constant, cannot be reassigned
+
+var a = 1, b = 2, c = 3  // multiple declarations
+```
+
+### Data Types
+
+ƿit has eight fundamental types:
+
+- **number** — DEC64 decimal floating point (no rounding errors)
+- **text** — Unicode strings
+- **logical** — `true` or `false`
+- **null** — the absence of a value (no `undefined`)
+- **array** — ordered, numerically-indexed sequences
+- **object** — key-value records with prototype inheritance
+- **blob** — binary data (bits, not bytes)
+- **function** — first-class callable values
+
+### Literals
+
+```javascript
+// Numbers
+42
+3.14
+-5
+0
+1e3           // scientific notation (1000)
+
+// Text
+"hello"
+`template ${x}`   // string interpolation
+`${1 + 2}`        // expression interpolation
+
+// Logical
+true
+false
+
+// Null
+null
+
+// Arrays
+[1, 2, 3]
+[]
+
+// Objects
+{a: 1, b: "two"}
+{}
+
+// Regex
+/\d+/
+/hello/i          // with flags
+```
+
+## Operators
+
+### Arithmetic
+
+```javascript
+2 + 3      // 5
+5 - 3      // 2
+3 * 4      // 12
+12 / 4     // 3
+10 % 3     // 1
+2 ** 3     // 8 (exponentiation)
+```
+
+### Comparison
+
+All comparisons are strict — there is no type coercion.
+
+```javascript
+5 == 5     // true
+5 != 6     // true
+3 < 5      // true
+5 > 3      // true
+3 <= 3     // true
+5 >= 5     // true
+```
+
+### Logical
+
+```javascript
+true && true     // true
+true && false    // false
+false || true    // true
+false || false   // false
+!true            // false
+!false           // true
+```
+
+Logical operators short-circuit:
+
+```javascript
+var called = false
+var fn = function() { called = true; return true }
+var r = false && fn()   // fn() not called
+r = true || fn()        // fn() not called
+```
+
+### Bitwise
+
+```javascript
+5 & 3       // 1   (AND)
+5 | 3       // 7   (OR)
+5 ^ 3       // 6   (XOR)
+~0          // -1  (NOT)
+1 << 3      // 8   (left shift)
+8 >> 3      // 1   (right shift)
+-1 >>> 1    // 2147483647 (unsigned right shift)
+```
+
+### Unary
+
+```javascript
++5          // 5
+-5          // -5
+-(-5)       // 5
+```
+
+### Increment and Decrement
+
+```javascript
+var x = 5
+x++         // returns 5, x becomes 6 (postfix)
+++x         // returns 7, x becomes 7 (prefix)
+x--         // returns 7, x becomes 6 (postfix)
+--x         // returns 5, x becomes 5 (prefix)
+```
+
+### Compound Assignment
+
+```javascript
+var x = 10
+x += 3      // 13
+x -= 3      // 10
+x *= 2      // 20
+x /= 4     // 5
+x %= 3      // 2
+```
+
+### Ternary
+
+```javascript
+var a = true ? 1 : 2       // 1
+var b = false ? 1 : 2      // 2
+var c = true ? (false ? 1 : 2) : 3  // 2 (nested)
+```
+
+### Comma
+
+The comma operator evaluates all expressions and returns the last.
+
+```javascript
+var x = (1, 2, 3)  // 3
+```
+
+### In
+
+Test whether a key exists in an object.
+
+```javascript
+var o = {a: 1}
+"a" in o    // true
+"b" in o    // false
+```
+
+### Delete
+
+Remove a key from an object.
+
+```javascript
+var o = {a: 1, b: 2}
+delete o.a
+"a" in o    // false
+o.b         // 2
+```
+
+## Property Access
+
+### Dot and Bracket
+
+```javascript
+var o = {x: 10}
+o.x           // 10 (dot read)
+o.x = 20      // dot write
+o["x"]        // 20 (bracket read)
+var key = "x"
+o[key]        // 20 (computed bracket)
+o["y"] = 30   // bracket write
+```
+
+### Object as Key
+
+Objects can be used as keys in other objects.
+
+```javascript
+var k = {}
+var o = {}
+o[k] = 42
+o[k]          // 42
+o[{}]         // null (different object)
+k in o        // true
+delete o[k]
+k in o        // false
+```
+
+### Chained Access
+
+```javascript
+var d = {a: {b: [1, {c: 99}]}}
+d.a.b[1].c    // 99
+```
+
+## Arrays
+
+Arrays are **distinct from objects**. They are ordered, numerically-indexed sequences.
+
+```javascript
+var arr = [1, 2, 3]
+arr[0]              // 1
+arr[2] = 10         // [1, 2, 10]
+length(arr)         // 3
+```
+
+### Push and Pop
+
+```javascript
+var a = [1, 2]
+a[] = 3             // push: [1, 2, 3]
+length(a)           // 3
+var v = a[]         // pop: v is 3, a is [1, 2]
+length(a)           // 2
+```
+
+## Objects
+
+Objects are key-value records with prototype-based inheritance.
+
+```javascript
+var point = {x: 10, y: 20}
+point.x             // 10
+point["y"]          // 20
+```
+
+### Prototypes
+
+```javascript
+// Create object with prototype
+var parent = {x: 10}
+var child = meme(parent)
+child.x             // 10 (inherited)
+proto(child)        // parent
+
+// Override does not mutate parent
+child.x = 20
+parent.x            // 10
+```
+
+### Mixins
+
+```javascript
+var p = {a: 1}
+var m1 = {b: 2}
+var m2 = {c: 3}
+var child = meme(p, [m1, m2])
+child.a             // 1 (from prototype)
+child.b             // 2 (from mixin)
+child.c             // 3 (from mixin)
+```
+
+## Control Flow
+
+### If / Else
+
+```javascript
+var x = 0
+if (true) x = 1
+if (false) x = 2 else x = 3
+if (false) x = 4
+else if (true) x = 5
+else x = 6
+```
+
+### While
+
+```javascript
+var i = 0
+while (i < 5) i++
+
+// break
+i = 0
+while (true) {
+  if (i >= 3) break
+  i++
+}
+
+// continue
+var sum = 0
+i = 0
+while (i < 5) {
+  i++
+  if (i % 2 == 0) continue
+  sum += i
+}
+```
+
+### For
+
+Variables cannot be declared in the for initializer. Declare them at the function body level.
+
+```javascript
+var sum = 0
+var i = 0
+for (i = 0; i < 5; i++) sum += i
+
+// break
+sum = 0
+i = 0
+for (i = 0; i < 10; i++) {
+  if (i == 5) break
+  sum += i
+}
+
+// continue
+sum = 0
+i = 0
+for (i = 0; i < 5; i++) {
+  if (i % 2 == 0) continue
+  sum += i
+}
+
+// nested
+sum = 0
+var j = 0
+for (i = 0; i < 3; i++) {
+  for (j = 0; j < 3; j++) {
+    sum++
+  }
+}
+```
+
+## Functions
+
+### Function Expressions
+
+```javascript
+var add = function(a, b) { return a + b }
+add(2, 3)           // 5
+```
+
+### Arrow Functions
+
+```javascript
+var double = x => x * 2
+double(5)           // 10
+
+var sum = (a, b) => a + b
+sum(2, 3)           // 5
+
+var block = x => {
+  var y = x * 2
+  return y + 1
+}
+block(5)            // 11
+```
+
+### Return
+
+A function with no `return` returns `null`. An early `return` exits immediately.
+
+```javascript
+var fn = function() { var x = 1 }
+fn()                // null
+
+var fn2 = function() { return 1; return 2 }
+fn2()               // 1
+```
+
+### Arguments
+
+Extra arguments are ignored. Missing arguments are `null`.
+
+```javascript
+var fn = function(a, b) { return a + b }
+fn(1, 2, 3)        // 3 (extra arg ignored)
+
+var fn2 = function(a, b) { return a }
+fn2(1)              // 1 (b is null)
+```
+
+### Immediately Invoked Function Expression
+
+```javascript
+var r = (function(x) { return x * 2 })(21)  // 42
+```
+
+### Closures
+
+Functions capture variables from their enclosing scope.
+
+```javascript
+var make = function(x) {
+  return function(y) { return x + y }
+}
+var add5 = make(5)
+add5(3)             // 8
+```
+
+Captured variables can be mutated:
+
+```javascript
+var counter = function() {
+  var n = 0
+  return function() { n = n + 1; return n }
+}
+var c = counter()
+c()                 // 1
+c()                 // 2
+```
+
+### Recursion
+
+```javascript
+var fact = function(n) {
+  if (n <= 1) return 1
+  return n * fact(n - 1)
+}
+fact(5)             // 120
+```
+
+### This Binding
+
+When a function is called as a method, `this` refers to the object.
+
+```javascript
+var obj = {
+  val: 10,
+  get: function() { return this.val }
+}
+obj.get()           // 10
+```
+
+### Currying
+
+```javascript
+var f = function(a) {
+  return function(b) {
+    return function(c) { return a + b + c }
+  }
+}
+f(1)(2)(3)          // 6
+```
+
+## Identifiers
+
+Identifiers can contain `?` and `!` characters, both as suffixes and mid-name.
+
+```javascript
+var nil? = (x) => x == null
+nil?(null)          // true
+nil?(42)            // false
+
+var set! = (x) => x + 1
+set!(5)             // 6
+
+var is?valid = (x) => x > 0
+is?valid(3)         // true
+
+var do!stuff = () => 42
+do!stuff()          // 42
+```
+
+The `?` in an identifier is not confused with the ternary operator:
+
+```javascript
+var nil? = (x) => x == null
+var a = nil?(null) ? "yes" : "no"   // "yes"
+```
+
+## Type Checking
+
+### Type Functions
+
+```javascript
+is_number(42)           // true
+is_text("hi")           // true
+is_logical(true)        // true
+is_object({})           // true
+is_array([])            // true
+is_function(function(){})  // true
+is_null(null)           // true
+is_object([])           // false (array is not object)
+is_array({})            // false (object is not array)
+```
+
+### Truthiness
+
+Falsy values: `false`, `0`, `""`, `null`. Everything else is truthy.
+
+```javascript
+if (0) ...              // not entered
+if ("") ...             // not entered
+if (null) ...           // not entered
+if (1) ...              // entered
+if ("hi") ...           // entered
+if ({}) ...             // entered
+if ([]) ...             // entered
+```
+
+## Immutability with Stone
+
+The `stone()` function makes values permanently immutable.
+
+```javascript
+var o = {x: 1}
+is_stone(o)             // false
+stone(o)
+is_stone(o)             // true
+o.x = 2                 // disrupts!
+```
+
+Stone is **deep** — all nested objects and arrays are also frozen. This cannot be reversed.
+
+## Function Proxy
+
+A function with two parameters (`name`, `args`) acts as a proxy when properties are accessed on it. Any method call on the function dispatches through the proxy.
+
+```javascript
+var proxy = function(name, args) {
+  return `${name}:${length(args)}`
+}
+proxy.hello()           // "hello:0"
+proxy.add(1, 2)         // "add:2"
+proxy["method"]()       // "method:0"
+var m = "dynamic"
+proxy[m]()              // "dynamic:0"
+```
+
+For non-proxy functions, property access disrupts:
+
+```javascript
+var fn = function() { return 1 }
+fn.foo                  // disrupts
+fn.foo = 1              // disrupts
+```
+
+## Regex
+
+Regex literals are written with forward slashes, with optional flags.
+
+```javascript
+var r = /\d+/
+var result = extract("abc123", r)
+result[0]               // "123"
+
+var ri = /hello/i
+var result2 = extract("Hello", ri)
+result2[0]              // "Hello"
+```
+
+## Error Handling
+
+ƿit uses `disrupt` and `disruption` for error handling. A `disrupt` signals that something went wrong. The `disruption` block attached to a function catches it.
+
+```javascript
+var safe_divide = function(a, b) {
+  if (b == 0) disrupt
+  return a / b
+} disruption {
+  print("something went wrong")
+}
+```
+
+`disrupt` is a bare keyword — it does not carry a value. The `disruption` block knows that something went wrong, but not what.
+
+### Re-raising
+
+A `disruption` block can re-raise by calling `disrupt` again:
+
+```javascript
+var outer = function() {
+  var inner = function() { disrupt } disruption { disrupt }
+  inner()
+} disruption {
+  // caught here after re-raise
+}
+outer()
+```
+
+### Testing for Disruption
+
+```javascript
+var should_disrupt = function(fn) {
+  var caught = false
+  var wrapper = function() {
+    fn()
+  } disruption {
+    caught = true
+  }
+  wrapper()
+  return caught
+}
+```
+
+If an actor has an unhandled disruption, it crashes.
+
+## Self-Referencing Structures
+
+Objects can reference themselves:
+
+```javascript
+var o = {name: "root"}
+o.self = o
+o.self.self.name        // "root"
+```
+
+## Variable Shadowing
+
+Inner functions can shadow outer variables:
+
+```javascript
+var x = 10
+var fn = function() {
+  var x = 20
+  return x
+}
+fn()                    // 20
+x                       // 10
+```

docs/library/.pages

@@ -1,10 +0,0 @@
-nav:
-  - text.md
-  - number.md
-  - array.md
-  - object.md
-  - blob.md
-  - time.md
-  - math.md
-  - json.md
-  - random.md

docs/library/_index.md

@@ -0,0 +1,18 @@
+---
+title: "Standard Library"
+description: "ƿit standard library modules"
+weight: 90
+type: "docs"
+---
+
+The standard library provides modules loaded with `use()`.
+
+| Module | Description |
+|--------|-------------|
+| [blob](/docs/library/blob/) | Binary data (bits, not bytes) |
+| [time](/docs/library/time/) | Time constants and conversions |
+| [math](/docs/library/math/) | Trigonometry, logarithms, roots |
+| [json](/docs/library/json/) | JSON encoding and decoding |
+| [random](/docs/library/random/) | Random number generation |
+
+The `text`, `number`, `array`, and `object` functions are intrinsics — they are always available without `use`. See [Built-in Functions](/docs/functions/) for the full list, and the individual reference pages for [text](/docs/library/text/), [number](/docs/library/number/), [array](/docs/library/array/), and [object](/docs/library/object/).

docs/library/array.md

@@ -1,12 +1,19 @@
-# array
+---
+title: "array"
+description: "Array creation and manipulation"
+weight: 30
+type: "docs"
+---
 
-The `array` function and its methods handle array creation and manipulation.
+The `array` function is an intrinsic (always available, no `use()` needed). It is **polymorphic** — its behavior depends on the type of the first argument.
 
-## Creation
+## From a Number
+
+Create an array of a given size.
 
 ### array(number)
 
-Create an array of specified size, filled with `null`.
+All elements initialized to `null`.
 
 ```javascript
 array(3)  // [null, null, null]
@@ -14,24 +21,36 @@ array(3)  // [null, null, null]
 
 ### array(number, initial)
 
-Create an array with initial values.
+All elements initialized to a value. If initial is a function, it is called for each element (passed the index if arity >= 1).
 
 ```javascript
 array(3, 0)           // [0, 0, 0]
 array(3, i => i * 2)  // [0, 2, 4]
 ```
 
+## From an Array
+
+Copy, map, concat, or slice.
+
 ### array(array)
 
-Copy an array.
+Copy an array (mutable).
 
 ```javascript
 var copy = array(original)
 ```
 
+### array(array, function)
+
+Map — call function with each element, collect results.
+
+```javascript
+array([1, 2, 3], x => x * 2)  // [2, 4, 6]
+```
+
 ### array(array, from, to)
 
-Slice an array.
+Slice — extract a sub-array. Negative indices count from end.
 
 ```javascript
 array([1, 2, 3, 4, 5], 1, 4)  // [2, 3, 4]
@@ -40,32 +59,36 @@ array([1, 2, 3], -2)          // [2, 3]
 
 ### array(array, another)
 
-Concatenate arrays.
+Concatenate two arrays.
 
 ```javascript
 array([1, 2], [3, 4])  // [1, 2, 3, 4]
 ```
 
-### array(object)
+## From a Record
 
-Get keys of an object.
+### array(record)
+
+Get the keys of a record as an array of text.
 
 ```javascript
 array({a: 1, b: 2})  // ["a", "b"]
 ```
 
+## From Text
+
 ### array(text)
 
-Split text into grapheme clusters.
+Split text into individual characters (grapheme clusters). This is the standard way to iterate over characters in a string.
 
 ```javascript
-array("hello")       // ["h", "e", "l", "l", "o"]
-array("👨‍👩‍👧")          // ["👨‍👩‍👧"]
+array("hello")  // ["h", "e", "l", "l", "o"]
+array("ƿit")    // ["ƿ", "i", "t"]
 ```
 
 ### array(text, separator)
 
-Split text by separator.
+Split text by a separator string.
 
 ```javascript
 array("a,b,c", ",")  // ["a", "b", "c"]
@@ -73,7 +96,7 @@ array("a,b,c", ",")  // ["a", "b", "c"]
 
 ### array(text, length)
 
-Split text into chunks.
+Dice text into chunks of a given length.
 
 ```javascript
 array("abcdef", 2)  // ["ab", "cd", "ef"]
@@ -87,13 +110,13 @@ Iterate over elements.
 
 ```javascript
 array.for([1, 2, 3], function(el, i) {
-  log.console(i, el)
+  print(i, el)
 })
 
 // With early exit
 array.for([1, 2, 3, 4], function(el) {
   if (el > 2) return true
-  log.console(el)
+  print(el)
 }, false, true)  // prints 1, 2
 ```
 

docs/library/blob.md

@@ -1,4 +1,9 @@
-# blob
+---
+title: "blob"
+description: "Binary data containers (bits, not bytes)"
+weight: 50
+type: "docs"
+---
 
 Blobs are binary large objects — containers of bits (not bytes). They're used for encoding data, messages, images, network payloads, and more.
 

docs/library/json.md

@@ -1,4 +1,9 @@
-# json
+---
+title: "json"
+description: "JSON encoding and decoding"
+weight: 80
+type: "docs"
+---
 
 JSON encoding and decoding.
 
@@ -86,5 +91,5 @@ var config_text = json.encode(config, 2)
 
 // Load configuration
 var loaded = json.decode(config_text)
-log.console(loaded.debug)  // true
+print(loaded.debug)  // true
 ```

docs/library/math.md

@@ -1,10 +1,15 @@
-# math
+---
+title: "math"
+description: "Trigonometry, logarithms, and roots"
+weight: 70
+type: "docs"
+---
 
-Cell provides three math modules with identical functions but different angle representations:
+ƿit provides three math modules with identical functions but different angle representations:
 
 ```javascript
 var math = use('math/radians')  // angles in radians
-var math = use('math/degrees')  // angles in degrees  
+var math = use('math/degrees')  // angles in degrees
 var math = use('math/cycles')   // angles in cycles (0-1)
 ```
 
@@ -35,7 +40,7 @@ math.tangent(math.pi / 4)  // 1 (radians)
 Inverse sine.
 
 ```javascript
-math.arc_sine(1)  // π/2 (radians)
+math.arc_sine(1)  // pi/2 (radians)
 ```
 
 ### arc_cosine(n)
@@ -43,7 +48,7 @@ math.arc_sine(1)  // π/2 (radians)
 Inverse cosine.
 
 ```javascript
-math.arc_cosine(0)  // π/2 (radians)
+math.arc_cosine(0)  // pi/2 (radians)
 ```
 
 ### arc_tangent(n, denominator)
@@ -51,9 +56,9 @@ math.arc_cosine(0)  // π/2 (radians)
 Inverse tangent. With two arguments, computes atan2.
 
 ```javascript
-math.arc_tangent(1)        // π/4 (radians)
-math.arc_tangent(1, 1)     // π/4 (radians)
-math.arc_tangent(-1, -1)   // -3π/4 (radians)
+math.arc_tangent(1)        // pi/4 (radians)
+math.arc_tangent(1, 1)     // pi/4 (radians)
+math.arc_tangent(-1, -1)   // -3pi/4 (radians)
 ```
 
 ## Exponentials and Logarithms
@@ -64,7 +69,7 @@ Euler's number raised to a power. Default power is 1.
 
 ```javascript
 math.e()     // 2.718281828...
-math.e(2)    // e²
+math.e(2)    // e^2
 ```
 
 ### ln(n)
@@ -130,21 +135,21 @@ math.e()  // 2.71828...
 var math = use('math/radians')
 
 // Distance between two points
-function distance(x1, y1, x2, y2) {
+var distance = function(x1, y1, x2, y2) {
   var dx = x2 - x1
   var dy = y2 - y1
   return math.sqrt(dx * dx + dy * dy)
 }
 
 // Angle between two points
-function angle(x1, y1, x2, y2) {
+var angle = function(x1, y1, x2, y2) {
   return math.arc_tangent(y2 - y1, x2 - x1)
 }
 
 // Rotate a point
-function rotate(x, y, angle) {
-  var c = math.cosine(angle)
-  var s = math.sine(angle)
+var rotate = function(x, y, a) {
+  var c = math.cosine(a)
+  var s = math.sine(a)
   return {
     x: x * c - y * s,
     y: x * s + y * c

docs/library/number.md

@@ -1,6 +1,11 @@
-# number
+---
+title: "number"
+description: "Numeric conversion and operations"
+weight: 20
+type: "docs"
+---
 
-The `number` function and its methods handle numeric conversion and operations.
+The `number` function is an intrinsic (always available, no `use()` needed). It is **polymorphic** — its behavior depends on the type of the first argument.
 
 ## Conversion
 
@@ -29,15 +34,15 @@ Parse formatted numbers.
 
 | Format | Description |
 |--------|-------------|
-| `""`   | Standard decimal |
-| `"u"`  | Underbar separator (1_000) |
-| `"d"`  | Comma separator (1,000) |
-| `"s"`  | Space separator (1 000) |
-| `"v"`  | European (1.000,50) |
-| `"b"`  | Binary |
-| `"o"`  | Octal |
-| `"h"`  | Hexadecimal |
-| `"j"`  | JavaScript style (0x, 0o, 0b prefixes) |
+| `""` | Standard decimal |
+| `"u"` | Underbar separator (1_000) |
+| `"d"` | Comma separator (1,000) |
+| `"s"` | Space separator (1 000) |
+| `"v"` | European (1.000,50) |
+| `"b"` | Binary |
+| `"o"` | Octal |
+| `"h"` | Hexadecimal |
+| `"j"` | JavaScript style (0x, 0o, 0b prefixes) |
 
 ```javascript
 number("1,000", "d")    // 1000
@@ -118,20 +123,20 @@ Get the fractional part.
 fraction(4.75)  // 0.75
 ```
 
-### min(...values)
+### min(a, b)
 
-Return the smallest value.
+Return the smaller of two numbers.
 
 ```javascript
-min(3, 1, 4, 1, 5)  // 1
+min(3, 5)  // 3
 ```
 
-### max(...values)
+### max(a, b)
 
-Return the largest value.
+Return the larger of two numbers.
 
 ```javascript
-max(3, 1, 4, 1, 5)  // 5
+max(3, 5)  // 5
 ```
 
 ### remainder(dividend, divisor)

docs/library/object.md

@@ -1,8 +1,13 @@
-# object
+---
+title: "object"
+description: "Object creation and manipulation"
+weight: 40
+type: "docs"
+---
 
-The `object` function and related utilities handle object creation and manipulation.
+The `object` function is an intrinsic (always available, no `use()` needed). It is **polymorphic** — its behavior depends on the types of its arguments.
 
-## Creation
+## From a Record
 
 ### object(obj)
 
@@ -29,6 +34,8 @@ Select specific keys.
 object({a: 1, b: 2, c: 3}, ["a", "c"])  // {a: 1, c: 3}
 ```
 
+## From an Array of Keys
+
 ### object(keys)
 
 Create object from keys (values are `true`).
@@ -60,9 +67,9 @@ object(["a", "b", "c"], (k, i) => i)  // {a: 0, b: 1, c: 2}
 Create a new object with the given prototype.
 
 ```javascript
-var animal = {speak: function() { log.console("...") }}
+var animal = {speak: function() { print("...") }}
 var dog = meme(animal)
-dog.speak = function() { log.console("woof") }
+dog.speak = function() { print("woof") }
 ```
 
 ### proto(obj)
@@ -104,9 +111,4 @@ var obj = {a: 1, b: 2, c: 3}
 
 // Get all keys
 var keys = array(obj)  // ["a", "b", "c"]
-
-// Iterate
-for (var key in obj) {
-  log.console(key, obj[key])
-}
 ```

docs/library/random.md

@@ -1,4 +1,9 @@
-# random
+---
+title: "random"
+description: "Random number generation"
+weight: 90
+type: "docs"
+---
 
 Random number generation.
 
@@ -43,7 +48,7 @@ var random = use('random')
 var coin_flip = random.random() < 0.5
 
 // Random element from array
-function pick(arr) {
+var pick = function(arr) {
   return arr[random.random_whole(length(arr))]
 }
 
@@ -51,11 +56,14 @@ var colors = ["red", "green", "blue"]
 var color = pick(colors)
 
 // Shuffle array
-function shuffle(arr) {
+var shuffle = function(arr) {
   var result = array(arr)  // copy
-  for (var i = length(result) - 1; i > 0; i--) {
-    var j = random.random_whole(i + 1)
-    var temp = result[i]
+  var i = length(result) - 1
+  var j = 0
+  var temp = null
+  for (i = length(result) - 1; i > 0; i--) {
+    j = random.random_whole(i + 1)
+    temp = result[i]
     result[i] = result[j]
     result[j] = temp
   }
@@ -63,8 +71,8 @@ function shuffle(arr) {
 }
 
 // Random in range
-function random_range(min, max) {
-  return min + random.random() * (max - min)
+var random_range = function(lo, hi) {
+  return lo + random.random() * (hi - lo)
 }
 
 var x = random_range(-10, 10)  // -10 to 10

docs/library/text.md

@@ -1,19 +1,28 @@
-# text
+---
+title: "text"
+description: "String conversion and manipulation"
+weight: 10
+type: "docs"
+---
 
-The `text` function and its methods handle string conversion and manipulation.
+The `text` function is an intrinsic (always available, no `use()` needed). It is **polymorphic** — its behavior depends on the type of the first argument.
 
-## Conversion
+To split text into characters, use `array(text)` — see [array](/docs/library/array/).
+
+## From an Array
 
 ### text(array, separator)
 
-Convert an array to text, joining elements with a separator (default: space).
+Join array elements into text with a separator (default: empty string).
 
 ```javascript
-text([1, 2, 3])           // "1 2 3"
-text([1, 2, 3], ", ")     // "1, 2, 3"
-text(["a", "b"], "-")     // "a-b"
+text(["h", "e", "l", "l", "o"])  // "hello"
+text([1, 2, 3], ", ")            // "1, 2, 3"
+text(["a", "b"], "-")            // "a-b"
 ```
 
+## From a Number
+
 ### text(number, radix)
 
 Convert a number to text. Radix is 2-36 (default: 10).
@@ -24,13 +33,16 @@ text(255, 16)    // "ff"
 text(255, 2)     // "11111111"
 ```
 
+## From Text
+
 ### text(text, from, to)
 
-Extract a substring from index `from` to `to`.
+Extract a substring from index `from` to `to`. Negative indices count from end.
 
 ```javascript
 text("hello world", 0, 5)   // "hello"
 text("hello world", 6)      // "world"
+text("hello", -3)            // "llo"
 ```
 
 ## Methods
@@ -101,7 +113,7 @@ text.format("{0} + {1} = {2}", [1, 2, 3])
 Unicode normalize the text (NFC form).
 
 ```javascript
-text.normalize("café")  // normalized form
+text.normalize("cafe\u0301")  // normalized form
 ```
 
 ### text.codepoint(text)
@@ -109,8 +121,7 @@ text.normalize("café")  // normalized form
 Get the Unicode codepoint of the first character.
 
 ```javascript
-text.codepoint("A")     // 65
-text.codepoint("😀")    // 128512
+text.codepoint("A")  // 65
 ```
 
 ### text.extract(text, pattern, from, to)

docs/library/time.md

@@ -1,4 +1,9 @@
-# time
+---
+title: "time"
+description: "Time constants and conversion functions"
+weight: 60
+type: "docs"
+---
 
 The time module provides time constants and conversion functions.
 
@@ -96,7 +101,7 @@ var last_week = now - time.week
 var later = now + (2 * time.hour)
 
 // Format future time
-log.console(time.text(tomorrow))
+print(time.text(tomorrow))
 ```
 
 ## Example
@@ -108,9 +113,9 @@ var time = use('time')
 var start = time.number()
 // ... do work ...
 var elapsed = time.number() - start
-log.console(`Took ${elapsed} seconds`)
+print(`Took ${elapsed} seconds`)
 
 // Schedule for tomorrow
 var tomorrow = time.number() + time.day
-log.console(`Tomorrow: ${time.text(tomorrow, "yyyy-MM-dd")}`)
+print(`Tomorrow: ${time.text(tomorrow, "yyyy-MM-dd")}`)
 ```

docs/memory.md

@@ -1,248 +0,0 @@
-# Cell actor scripting language
-
-Cell is a Misty [https://mistysystem.com](https://mistysystem.com) implementation.
-
-## Memory
-Values are 32 bit for 32 bit builds and 64 bit for 64 bit builds.
-
-### 32 bit value
-
-LSB = 0
-payload is a 31 bit signed int
-
-LSB = 01
-payload is a 30 bit pointer
-
-LSB = 11
-next 3 bits = special tag. 27 bits of payload.
-
-### 64 bit value
-LSB = 0
-payload is a 32 bit signed int, using high 32 bits
-
-LSB = 01
-payload is a 61 bit pointer
-
-LSB = 101
-Short float: a 61 bit double, with 3 less exponent bits
-
-LSB = 11
-Special tag: next 3 bits. 5 bits total. 59 bits of payload. 8 total special tags.
-
-Special tags:
-1: Bool. Payload is 0 or 1.
-2: null. payload is 0.
-3: exception.
-4: string.
-  Immediate string. Next 3 low bits = length in bytes. Rest is string data. This allows for strings up to 7 ascii letters. Encoded in utf8.
-
-## Numbers and math
-Cell can be compiled with different levels of exactness for numeracy. Any number which cannot be represented exactly becomes "null". Any numeric operation which includes "null" results in "null".
-
-Using short floats in a 64 bit system means you have doubles in the range of +- 10^38, not the full range of double. If you create a number out of that range, it's null.
-
-You can also compile a 64 bit system with full precision doubles, but this will use more memory and may be slower.
-
-You can also compile a 64 bit system with 32 bit floats, stored as a 32 bit int is. Again, out of the 32 bit float range = null.
-
-You can compile without floating point support at all; 32 bit ints are then used for fixed point calculations.
-
-Or, you can compile using Dec64, which is a 64 bit decimal floating point format, for exact precision.
-
-## Objects
-Objects are heap allocated, referenced by a pointer value. They are all preceded by an object header, the length of a word on the system.
-  
-### 64 bit build
-56 bits capacity
-1 bit memory reclamation flag: note that this obj has already been moved
-2 bit reserved (per object)
-1 bit stone: note that this obj is immutable
-3 bit type: note the type of the object
-1 bit: fwd: note that this obj is a forward linkage
-
-Last bit ..1:
-The forward type indicates that the object (an array, blob, pretext, or record) has grown beyond its capacity and is now residing at a new address. The remaining 63 bits contain the address of the enlarged object. Forward linkages are cleaned up by the memory reclaimer.
-
-Type 7: C light C object
-
-Header
-Pointer
-
-Capacity is an ID of a registered C type.
-Pointer is a pointer to the opaque C object.
-
-Type 0: Array
-Header
-Length
-Element[]
-
-Capacity is number of elements the array can hold. Length is number of elements in use. Number of words used by an array is capacity + 2.
-
-Type 1: blob
-Header
-Length
-Bit[]
-Capacity is number of bits the blob can hold. Length is number of bits in use. Bits follow, from [0] to [capacity - 1], with [0] bit in the most significant position of word 2, and [63] in the least significant position of word 2. The last word is zero filled, if necessary.
-
-Number of words used is (capacity + 63) // 64 + 2
-
-Type 2: Text
-Text has two forms, depending on if it is stone or not, which changes the meaning of its length word.
-
-Header
-Length(pretext) or Hash(text)
-Character[0] and character[1]
-
-Capacity of pretex is the number of characters it can hold. During stoning and reclamation, capacity is set to the length.
-
-The capacity of a text is its length.
-
-The length of a pretext is the number of characters it contains; it is not greater than the capacity.
-
-Hash of a text is used for organizing records. If the hash is zero, it's not been computed yet. All texts in the immutable memory have hashes.
-
-A text object contains UTF32 characters, packed two per word. If the number of characters is odd, the least significant half of the last word is zero filled.
-
-The number of words used by a text is (capacity + 1) // 2 + 2
-
-Type 3: Record
-
-A record is an array of fields represented as key/value pairs. Fields are located by hashes of texts, using open addressing with linear probing and lazy deletion. The load factor is less than 0.5.
-
-Header
-Prototype
-Length
-Key[0]
-Value[0]
-Key[1]
-Value[1]
-...
-
-The capacity is the number of fields the record can hold. It is a power of two minus one. It is at least twice the length.
-
-The length is the number of fields that the record currently contains.
-
-A field candidate number is identified by and(key.hash, capacity). In case of hash collision, advance to the next field. If this goes past the end, continue with field 1. Field 0 is reserved.
-
-The "exception" special tag is used to mark deleted entries in the object map.
-
-The number of words used by a record is (capacity + 1) * 2.
-
-Prototypes are searched for for properties if one cannot be found on the record itself. Prototypes can have prototypes.
-
-#### key[0] and value[0]
-These are reserved for internal use, and skipped over during key probing.
-
-The first 32 bits of key are used as a 32 bit integer key, if this object has ever been used as a key itself.
-
-The last 32 bits are used as an opaque C class key. C types can be registered with the system, and each are assigned a monotonically increasing number. In the case that this object has a C type, then the bottom 32 bits of key[0] are not 0. If that is the case, then a pointer to its C object is stored in value[0].
-
-#### Valid keys & Hashing
-Keys are stored directly in object maps. There are three possibilities for a vaild key: an object text, an object record, or an immediate text.
-
-In the case of an immediate text, the hash is computed on the fly using the fash64_hash_one function, before being used to look up the key in the object map. Direct value comparison is used to confirm the key.
-
-For object texts (texts longer than 7 ascii chars), the hash is stored in the text object itself. When an object text is used as a key, a stone version is created and interned. Any program static texts reference this stoned, interned text. When looking up a heap text as a key, it is first discovered if it's in the interned table. If it's not, the key is not in the object (since all keys are interned). If it is, the interned version is returned to check against the object map. The hash of the interned text is used to look up the key in the object map, and then direct pointer comparison is used to confirm the key.
-
-For record keys, these are unique; once a record is used as a key, it gets assigned a monotonically increasing 32 bit integer, stored in key[0]. When checking it in an object map, the integer is used directly as the key. If key[0] is 0, the record has not been used as a key yet. If it's not 0, fash64_hash_one is used to compute a hash of its ID, and then direct value pointer comparison is used to confirm.
-
-### Text interning
-Texts that cannot fit in an immediate, and which are used as an object key, create a stoned and interned version (the pointer which is used as the key). Any text literals are also stoned and interned.
-
-The interning table is an open addressed hash, with a load of 0.8, using a robin hood value. Probing is done using the text hash, confirmation is done using length, and then memcmp of the text.
-
-When the GC run, a new interned text table is created. Each text literal, and each text used as a key, is added to the new table, as the live objects are copied. This keeps the interning table from becoming a graveyard. Interned values are never deleted until a GC.
-
-Type 4: Function
-
-Header
-Code
-Outer
-A function object has zero capacity and is always stone.
-
-Code is a pointer to the code object that the function executes.
-
-Outer is a pointer to the frame that created this function object.
-
-Size is 3 words.
-
-Type 5: Frame
-
-Header
-Function
-Caller
-Return address
-
-The activation frame is created when a function is invoked to hold its linkages and state.
-
-The capacity is the number of slots, including the inputs, variables, temporaries, and the four words of overhead. A frame, unlike the other types, is never stone.
-
-The function is the address of the function object being called.
-
-The caller is the address of the frame that is invoking the function.
-
-The return address is the address of the instruction in the code that should be executed upon return.
-
-Next come the input arguments, if any.
-
-Then the variables closed over by the inner functions.
-
-Then the variables that are not closed over, followed by the temporaries.
-
-When a function returns, the caller is set to zero. This is a signal to the memory reclaimer that the frame can be reduced.
-
-Type 6: Code
-
-Header
-Arity
-Size
-Closure size
-Entry point
-Disruption point
-
-A code object exists in the actor's immutable memory. A code object never exists in mutable memory.
-
-A code object has a zero capacity and is always stone.
-
-The arity is the maximum number of inputs.
-
-The size is the capacity of an activation frame that will execute this code.
-
-The closure size is a reduced capacity for returned frames that survive memory reclamation.
-
-The entry point is the address at which to begin execution.
-
-The disruption point is the address of the disruption clause.
-
-### opaque C objects
-Records can have opaque C data attached to them.
-
-A C class can register a GC clean up, and a GC trace function. The trace function is called when the record is encountered in the live object graph; and it should mark any values it wants to keep alive in that function.
-
-The system maintains an array of live opaque C objects. When such an object is encountered, it marks it as live in the array. When the GC completes, it iterates this array and calls the GC clean up function for each C object in the array with alive=0. Alive is then cleared for the next GC cycle.
-
-## 32 bit build
-~3 bit type
-1 bit stone
-1 bit memory reclamation flag
-27 bit capacity
-
-Key differences here are
-
-blob max capacity is 2**27 bits = 2**24 bytes = 16 MB [this likely needs addressed]
-
-fwd is type ...0, and the pointer is 31 bits
-other types are
-111 array
-101 object
-011 blob
-001 
-
-## Memory
-Cell uses a single block of memory that it doles out as needed to the actors in its system.
-
-Actors are given a block of memory in standard sizes using a doubling buddy memory manager. An actor is given an immutable data section on birth, as well as a mutable data section. When its mutable data becomes full, it requests a new one. Actors utilize their mutable memory with a simple bump allocation. If there is not sufficient memory available, the actor suspends and its status changes to exhausted.
-
-The smallest block size is determined per platform, but it can be as small as 4KB on 64 bit systems.
-
-The actor is then given a new block of memory of the same size, and it runs a garbage collector to reclaim memory. It uses the cheney copying algorithm. If a disappointing amount of memory was reclaimed, it is noted, and the actor is given a larger block of memory on the next request.

docs/nota.md

@@ -0,0 +1,156 @@
+---
+title: "Nota Format"
+description: "Network Object Transfer Arrangement"
+weight: 85
+type: "docs"
+---
+
+Nota is a binary message format developed for use in the Procession Protocol. It provides a compact, JSON-like encoding that supports blobs, text, arrays, records, numbers, and symbols.
+
+Nota stands for Network Object Transfer Arrangement.
+
+## Design Philosophy
+
+JSON had three design rules: minimal, textual, and subset of JavaScript. The textual and JavaScript rules are no longer necessary. Nota maintains JSON's philosophy of being at the intersection of most programming languages and most data types, but departs by using counts instead of brackets and binary encoding instead of text.
+
+Nota uses Kim continuation bytes for counts and character encoding. See [Kim Encoding](#kim) for details.
+
+## Type Summary
+
+| Bits | Type |
+|------|------|
+| `000` | Blob |
+| `001` | Text |
+| `010` | Array |
+| `011` | Record |
+| `100` | Floating Point (positive exponent) |
+| `101` | Floating Point (negative exponent) |
+| `110` | Integer (zero exponent) |
+| `111` | Symbol |
+
+## Preambles
+
+Every Nota value starts with a preamble byte that is a Kim value with the three most significant bits used for type information.
+
+Most types provide 3 or 4 data bits in the preamble. If the Kim encoding of the data fits in those bits, it is incorporated directly and the continue bit is off. Otherwise the continue bit is on and the continuation follows.
+
+## Blob
+
+```
+C  0  0  0  D  D  D  D
+```
+
+- **C** — continue the number of bits
+- **DDDD** — the number of bits
+
+A blob is a string of bits. The data produces the number of bits. The number of bytes that follow: `floor((number_of_bits + 7) / 8)`. The final byte is padded with 0 if necessary.
+
+Example: A blob containing 25 bits `1111000011100011001000001`:
+
+```
+80 19 F0 E3 20 80
+```
+
+## Text
+
+```
+C  0  0  1  D  D  D  D
+```
+
+- **C** — continue the number of characters
+- **DDDD** — the number of characters
+
+The data produces the number of characters. Kim-encoded characters follow. ASCII characters are 1 byte, first quarter BMP characters are 2 bytes, all other Unicode characters are 3 bytes. Unlike JSON, there is never a need for escapement.
+
+Examples:
+
+```
+""          →  10
+"cat"       →  13 63 61 74
+```
+
+## Array
+
+```
+C  0  1  0  D  D  D  D
+```
+
+- **C** — continue the number of elements
+- **DDDD** — the number of elements
+
+An array is an ordered sequence of values. Following the preamble are the elements, each beginning with its own preamble. Nesting is encouraged.
+
+## Record
+
+```
+C  0  1  1  D  D  D  D
+```
+
+- **C** — continue the number of pairs
+- **DDDD** — the number of pairs
+
+A record is an unordered collection of key/value pairs. Keys must be text and must be unique within the record. Values can be any Nota type.
+
+## Floating Point
+
+```
+C  1  0  E  S  D  D  D
+```
+
+- **C** — continue the exponent
+- **E** — sign of the exponent
+- **S** — sign of the coefficient
+- **DDD** — three bits of the exponent
+
+Nota floating point represents numbers as `coefficient * 10^exponent`. The coefficient must be an integer. The preamble may contain the first three bits of the exponent, followed by the continuation of the exponent (if any), followed by the coefficient.
+
+Use the integer type when the exponent is zero.
+
+Examples:
+
+```
+-1.01                →  5A 65
+98.6                 →  51 87 5A
+-0.5772156649        →  D8 0A 95 C0 B0 BD 69
+-10000000000000      →  C8 0D 01
+```
+
+## Integer
+
+```
+C  1  1  0  S  D  D  D
+```
+
+- **C** — continue the integer
+- **S** — sign
+- **DDD** — three bits of the integer
+
+Integers in the range -7 to 7 fit in a single byte. Integers in the range -1023 to 1023 fit in two bytes. Integers in the range -131071 to 131071 fit in three bytes.
+
+Examples:
+
+```
+0      →  60
+2023   →  E0 8F 67
+-1     →  69
+```
+
+## Symbol
+
+```
+0  1  1  1  D  D  D  D
+```
+
+- **DDDD** — the symbol
+
+There are currently five symbols:
+
+```
+null     →  70
+false    →  72
+true     →  73
+private  →  78
+system   →  79
+```
+
+The private prefix must be followed by a record containing a private process address. The system prefix must be followed by a record containing a system message. All other symbols are reserved.

docs/packages.md

@@ -1,14 +1,19 @@
-# Packages
+---
+title: "Packages"
+description: "Code organization and sharing in ƿit"
+weight: 30
+type: "docs"
+---
 
-Packages are the fundamental unit of code organization and sharing in Cell.
+Packages are the fundamental unit of code organization and sharing in ƿit.
 
 ## Package Structure
 
-A package is a directory containing a `cell.toml` manifest:
+A package is a directory containing a `pit.toml` manifest:
 
 ```
 mypackage/
-├── cell.toml        # package manifest
+├── pit.toml         # package manifest
 ├── main.ce          # entry point (optional)
 ├── utils.cm         # module
 ├── helper/
@@ -17,7 +22,7 @@ mypackage/
 └── _internal.cm     # private module (underscore prefix)
 ```
 
-## cell.toml
+## pit.toml
 
 The package manifest declares metadata and dependencies:
 
@@ -38,11 +43,11 @@ mylib = "/Users/john/work/mylib"
 
 ## Module Resolution
 
-When importing with `use()`, Cell searches in order:
+When importing with `use()`, ƿit searches in order:
 
 1. **Local package** — relative to package root
-2. **Dependencies** — via aliases in `cell.toml`
-3. **Core** — built-in Cell modules
+2. **Dependencies** — via aliases in `pit.toml`
+3. **Core** — built-in ƿit modules
 
 ```javascript
 // In package 'myapp' with dependency: renderer = "gitea.pockle.world/john/renderer"
@@ -85,10 +90,10 @@ Local packages are symlinked into the shop, making development seamless.
 
 ## The Shop
 
-Cell stores all packages in the **shop** at `~/.cell/`:
+ƿit stores all packages in the **shop** at `~/.pit/`:
 
 ```
-~/.cell/
+~/.pit/
 ├── packages/
 │   ├── core -> gitea.pockle.world/john/cell
 │   ├── gitea.pockle.world/
@@ -134,20 +139,20 @@ target = "/Users/john/work/prosperon"
 
 ```bash
 # Install from remote
-cell install gitea.pockle.world/john/prosperon
+pit install gitea.pockle.world/john/prosperon
 
 # Install from local path
-cell install /Users/john/work/mylib
+pit install /Users/john/work/mylib
 ```
 
 ## Updating Packages
 
 ```bash
 # Update all
-cell update
+pit update
 
 # Update specific package
-cell update gitea.pockle.world/john/prosperon
+pit update gitea.pockle.world/john/prosperon
 ```
 
 ## Development Workflow
@@ -156,12 +161,12 @@ For active development, link packages locally:
 
 ```bash
 # Link a package for development
-cell link add gitea.pockle.world/john/prosperon /Users/john/work/prosperon
+pit link add gitea.pockle.world/john/prosperon /Users/john/work/prosperon
 
 # Changes to /Users/john/work/prosperon are immediately visible
 
 # Remove link when done
-cell link delete gitea.pockle.world/john/prosperon
+pit link delete gitea.pockle.world/john/prosperon
 ```
 
 ## C Extensions
@@ -170,14 +175,14 @@ C files in a package are compiled into a dynamic library:
 
 ```
 mypackage/
-├── cell.toml
+├── pit.toml
 ├── render.c      # compiled to mypackage.dylib
-└── render.cm     # optional Cell wrapper
+└── render.cm     # optional ƿit wrapper
 ```
 
-The library is named after the package and placed in `~/.cell/lib/`.
+The library is named after the package and placed in `~/.pit/lib/`.
 
-See [Writing C Modules](c-modules.md) for details.
+See [Writing C Modules](/docs/c-modules/) for details.
 
 ## Platform-Specific Files
 
@@ -190,4 +195,4 @@ mypackage/
 └── audio_emscripten.c # Web-specific
 ```
 
-Cell selects the appropriate file based on the build target.
+ƿit selects the appropriate file based on the build target.

docs/requestors.md

@@ -0,0 +1,176 @@
+---
+title: "Requestors"
+description: "Asynchronous work with requestors"
+weight: 25
+type: "docs"
+---
+
+Requestors are functions that encapsulate asynchronous work. They provide a structured way to compose callbacks, manage cancellation, and coordinate concurrent operations between actors.
+
+## What is a Requestor
+
+A requestor is a function with this signature:
+
+```javascript
+var my_requestor = function(callback, value) {
+  // Do async work, then call callback with result
+  // Return a cancel function
+}
+```
+
+- **callback** — called when the work completes: `callback(value, reason)`
+  - On success: `callback(result)` or `callback(result, null)`
+  - On failure: `callback(null, reason)` where reason explains the failure
+- **value** — input passed from the previous step (or the initial caller)
+- **return** — a cancel function, or null if cancellation is not supported
+
+The cancel function, when called, should abort the in-progress work.
+
+## Writing a Requestor
+
+```javascript
+var fetch_data = function(callback, url) {
+  $contact(function(connection) {
+    $send(connection, {get: url}, function(response) {
+      callback(response)
+    })
+  }, {host: url, port: 80})
+  return function() {
+    // clean up if needed
+  }
+}
+```
+
+A requestor that always succeeds immediately:
+
+```javascript
+var constant = function(callback, value) {
+  callback(42)
+}
+```
+
+A requestor that always fails:
+
+```javascript
+var broken = function(callback, value) {
+  callback(null, "something went wrong")
+}
+```
+
+## Composing Requestors
+
+ƿit provides four built-in functions for composing requestors into pipelines.
+
+### sequence(requestor_array)
+
+Run requestors one after another. Each result becomes the input to the next. The final result is passed to the callback.
+
+```javascript
+var pipeline = sequence([
+  fetch_user,
+  validate_permissions,
+  load_profile
+])
+
+pipeline(function(profile, reason) {
+  if (reason) {
+    print(reason)
+  } else {
+    print(profile.name)
+  }
+}, user_id)
+```
+
+If any step fails, the remaining steps are skipped and the failure propagates.
+
+### parallel(requestor_array, throttle, need)
+
+Start all requestors concurrently. Results are collected into an array matching the input order.
+
+```javascript
+var both = parallel([
+  fetch_profile,
+  fetch_settings
+])
+
+both(function(results, reason) {
+  var profile = results[0]
+  var settings = results[1]
+}, user_id)
+```
+
+- **throttle** — limit how many requestors run at once (null for no limit)
+- **need** — minimum number of successes required (default: all)
+
+### race(requestor_array, throttle, need)
+
+Like `parallel`, but returns as soon as the needed number of results arrive. Unfinished requestors are cancelled.
+
+```javascript
+var fastest = race([
+  fetch_from_cache,
+  fetch_from_network,
+  fetch_from_backup
+])
+
+fastest(function(results) {
+  // results[0] is whichever responded first
+}, request)
+```
+
+Default need is 1. Useful for redundant operations where only one result matters.
+
+### fallback(requestor_array)
+
+Try each requestor in order. If one fails, try the next. Return the first success.
+
+```javascript
+var resilient = fallback([
+  fetch_from_primary,
+  fetch_from_secondary,
+  use_cached_value
+])
+
+resilient(function(data, reason) {
+  if (reason) {
+    print("all sources failed")
+  }
+}, key)
+```
+
+## Timeouts
+
+Wrap any requestor with `$time_limit` to add a timeout:
+
+```javascript
+var timed = $time_limit(fetch_data, 5)  // 5 second timeout
+
+timed(function(result, reason) {
+  // reason will explain timeout if it fires
+}, url)
+```
+
+If the requestor does not complete within the time limit, it is cancelled and the callback receives a failure.
+
+## Requestors and Actors
+
+Requestors are particularly useful with actor messaging. Since `$send` is callback-based, it fits naturally:
+
+```javascript
+var ask_worker = function(callback, task) {
+  $send(worker, task, function(reply) {
+    callback(reply)
+  })
+}
+
+var pipeline = sequence([
+  ask_worker,
+  process_result,
+  store_result
+])
+
+pipeline(function(stored) {
+  print("done")
+  $stop()
+}, {type: "compute", data: [1, 2, 3]})
+```

docs/spec/dec64.md

@@ -0,0 +1,77 @@
+---
+title: "DEC64 Numbers"
+description: "Decimal floating point representation"
+---
+
+## Overview
+
+ƿit uses DEC64 as its number format. DEC64 represents numbers as `coefficient * 10^exponent` in a 64-bit word. This eliminates the rounding errors that plague IEEE 754 binary floating point — `0.1 + 0.2` is exactly `0.3`.
+
+DEC64 was designed by Douglas Crockford as a general-purpose number type suitable for both business and scientific computation.
+
+## Format
+
+A DEC64 number is a 64-bit value:
+
+```
+[coefficient: 56 bits][exponent: 8 bits]
+```
+
+- **Coefficient** — a 56-bit signed integer (two's complement)
+- **Exponent** — an 8-bit signed integer (range: -127 to 127)
+
+The value of a DEC64 number is: `coefficient * 10^exponent`
+
+### Examples
+
+| Value | Coefficient | Exponent | Hex |
+|-------|------------|----------|-----|
+| `0` | 0 | 0 | `0000000000000000` |
+| `1` | 1 | 0 | `0000000000000100` |
+| `3.14159` | 314159 | -5 | `000000004CB2FFFB` |
+| `-1` | -1 | 0 | `FFFFFFFFFFFFFF00` |
+| `1000000` | 1 | 6 | `0000000000000106` |
+
+## Special Values
+
+### Null
+
+The exponent `0x80` (-128) indicates null. This is the only special value — there is no infinity, no NaN, no negative zero. Operations that would produce undefined results (such as division by zero) return null.
+
+```
+coefficient: any, exponent: 0x80  →  null
+```
+
+## Arithmetic Properties
+
+- **Exact decimals**: All decimal fractions with up to 17 significant digits are represented exactly
+- **No rounding**: `0.1 + 0.2 == 0.3` is true
+- **Integer range**: Exact integers up to 2^55 (about 3.6 * 10^16)
+- **Normalized on demand**: The runtime normalizes coefficients to remove trailing zeros when needed for comparison
+
+## Comparison with IEEE 754
+
+| Property | DEC64 | IEEE 754 double |
+|----------|-------|----------------|
+| Decimal fractions | Exact | Approximate |
+| Significant digits | ~17 | ~15-16 |
+| Special values | null only | NaN, ±Infinity, -0 |
+| Rounding errors | None (decimal) | Common |
+| Financial arithmetic | Correct | Requires libraries |
+| Scientific range | ±10^127 | ±10^308 |
+
+DEC64 trades a smaller exponent range for exact decimal arithmetic. Most applications never need exponents beyond ±127.
+
+## In ƿit
+
+All numbers in ƿit are DEC64. There is no separate integer type at the language level — the distinction is internal. The `is_integer` function checks whether a number has no fractional part.
+
+```javascript
+var x = 42        // coefficient: 42, exponent: 0
+var y = 3.14      // coefficient: 314, exponent: -2
+var z = 1000000   // coefficient: 1, exponent: 6 (normalized)
+
+is_integer(x)     // true
+is_integer(y)     // false
+1 / 0             // null
+```

docs/spec/gc.md

@@ -0,0 +1,82 @@
+---
+title: "Garbage Collection"
+description: "Cheney copying collector"
+---
+
+## Overview
+
+ƿit uses a Cheney copying collector for automatic memory management. Each actor has its own independent heap — actors never share mutable memory, so garbage collection is per-actor with no global pauses.
+
+## Algorithm
+
+The Cheney algorithm is a two-space copying collector:
+
+1. **Allocate new space** — a fresh memory block for the new heap
+2. **Copy roots** — copy all live root objects from old space to new space
+3. **Scan** — walk the new space, updating all internal references
+4. **Free old space** — the entire old heap is freed at once
+
+### Copying and Forwarding
+
+When an object is copied from old space to new space:
+
+1. The object's data is copied to the next free position in new space
+2. The old object's header is overwritten with a **forwarding pointer** (`OBJ_FORWARD`) containing the new address
+3. Future references to the old address find the forwarding pointer and follow it to the new location
+
+```
+Old space:                 New space:
+┌──────────────┐          ┌──────────────┐
+│ OBJ_FORWARD ─┼────────> │ copied object│
+│ (new addr)   │          │              │
+└──────────────┘          └──────────────┘
+```
+
+### Scan Phase
+
+After roots are copied, the collector scans new space linearly. For each object, it examines every JSValue field:
+
+- If the field points to old space, copy the referenced object (or follow its forwarding pointer if already copied)
+- If the field points to stone memory, skip it (stone objects are permanent)
+- If the field is an immediate value (integer, boolean, null, immediate string), skip it
+
+The scan continues until the scan pointer catches up with the allocation pointer — at that point, all live objects have been found and copied.
+
+## Roots
+
+The collector traces from these root sources:
+
+- **Global object** — all global variables
+- **Class prototypes** — built-in type prototypes
+- **Exception** — the current exception value
+- **Value stack** — all values on the operand stack
+- **Frame stack** — all stack frames (register VM and mcode)
+- **GC reference stack** — manually registered roots (via `JS_PUSH_VALUE` / `JS_POP_VALUE`)
+- **Parser constant pool** — during compilation, constants being built
+
+## Per-Actor Heaps
+
+Each actor maintains its own heap with independent collection:
+
+- No stop-the-world pauses across actors
+- No synchronization between collectors
+- Each actor's GC runs at the end of a turn (between message deliveries)
+- Heap sizes adapt independently based on each actor's allocation patterns
+
+## Heap Growth
+
+The collector uses a buddy allocator for heap blocks. After each collection, if less than 20% of the heap was recovered, the next block size is doubled. The new space size is: `max(live_estimate + alloc_size, next_block_size)`.
+
+All allocations within a heap block use bump allocation (advance a pointer), which is extremely fast.
+
+## Alignment
+
+All objects are aligned to 8-byte boundaries. Object sizes are rounded up to ensure this alignment, which guarantees that the low 3 bits of any heap pointer are always zero — available for JSValue tag bits.
+
+## Interaction with Stone Memory
+
+Stone memory objects (S bit set) are never copied by the collector. When the scanner encounters a pointer to stone memory, it leaves it unchanged. This means:
+
+- Stone objects are effectively permanent GC roots
+- No overhead for tracing through immutable object graphs
+- Module return values and interned strings impose zero GC cost

docs/spec/mach.md

@@ -0,0 +1,156 @@
+---
+title: "Register VM"
+description: "Register-based virtual machine (Mach)"
+---
+
+## Overview
+
+The Mach VM is a register-based virtual machine using 32-bit instructions. It is modeled after Lua's register VM — operands are register indices rather than stack positions, reducing instruction count and improving performance.
+
+## Instruction Formats
+
+All instructions are 32 bits wide. Four encoding formats are used:
+
+### iABC — Three-Register
+
+```
+[op: 8][A: 8][B: 8][C: 8]
+```
+
+Used for operations on three registers: `R(A) = R(B) op R(C)`.
+
+### iABx — Register + Constant
+
+```
+[op: 8][A: 8][Bx: 16]
+```
+
+Used for loading constants: `R(A) = K(Bx)`.
+
+### iAsBx — Register + Signed Offset
+
+```
+[op: 8][A: 8][sBx: 16]
+```
+
+Used for conditional jumps: if `R(A)` then jump by `sBx`.
+
+### isJ — Signed Jump
+
+```
+[op: 8][sJ: 24]
+```
+
+Used for unconditional jumps with a 24-bit signed offset.
+
+## Registers
+
+Each function frame has a fixed number of register slots, determined at compile time. Registers hold:
+
+- **R(0)** — `this` binding
+- **R(1)..R(arity)** — function arguments
+- **R(arity+1)..** — local variables and temporaries
+
+## Instruction Set
+
+### Loading
+
+| Opcode | Format | Description |
+|--------|--------|-------------|
+| `LOADK` | iABx | `R(A) = K(Bx)` — load from constant pool |
+| `LOADI` | iAsBx | `R(A) = sBx` — load small integer |
+| `LOADNULL` | iA | `R(A) = null` |
+| `LOADTRUE` | iA | `R(A) = true` |
+| `LOADFALSE` | iA | `R(A) = false` |
+| `MOVE` | iABC | `R(A) = R(B)` — register copy |
+
+### Arithmetic
+
+| Opcode | Format | Description |
+|--------|--------|-------------|
+| `ADD` | iABC | `R(A) = R(B) + R(C)` |
+| `SUB` | iABC | `R(A) = R(B) - R(C)` |
+| `MUL` | iABC | `R(A) = R(B) * R(C)` |
+| `DIV` | iABC | `R(A) = R(B) / R(C)` |
+| `MOD` | iABC | `R(A) = R(B) % R(C)` |
+| `POW` | iABC | `R(A) = R(B) ^ R(C)` |
+| `NEG` | iABC | `R(A) = -R(B)` |
+| `INC` | iABC | `R(A) = R(B) + 1` |
+| `DEC` | iABC | `R(A) = R(B) - 1` |
+
+### Comparison
+
+| Opcode | Format | Description |
+|--------|--------|-------------|
+| `EQ` | iABC | `R(A) = R(B) == R(C)` |
+| `NEQ` | iABC | `R(A) = R(B) != R(C)` |
+| `LT` | iABC | `R(A) = R(B) < R(C)` |
+| `LE` | iABC | `R(A) = R(B) <= R(C)` |
+| `GT` | iABC | `R(A) = R(B) > R(C)` |
+| `GE` | iABC | `R(A) = R(B) >= R(C)` |
+
+### Property Access
+
+| Opcode | Format | Description |
+|--------|--------|-------------|
+| `GETFIELD` | iABC | `R(A) = R(B)[K(C)]` — named property |
+| `SETFIELD` | iABC | `R(A)[K(B)] = R(C)` — set named property |
+| `GETINDEX` | iABC | `R(A) = R(B)[R(C)]` — computed property |
+| `SETINDEX` | iABC | `R(A)[R(B)] = R(C)` — set computed property |
+
+### Variable Resolution
+
+| Opcode | Format | Description |
+|--------|--------|-------------|
+| `GETNAME` | iABx | Unresolved variable (compiler placeholder) |
+| `GETINTRINSIC` | iABx | Global intrinsic / built-in |
+| `GETENV` | iABx | Module environment variable |
+| `GETUP` | iABC | `R(A) = UpFrame(B).slots[C]` — closure upvalue |
+| `SETUP` | iABC | `UpFrame(A).slots[B] = R(C)` — set closure upvalue |
+
+### Control Flow
+
+| Opcode | Format | Description |
+|--------|--------|-------------|
+| `JMP` | isJ | Unconditional jump |
+| `JMPTRUE` | iAsBx | Jump if `R(A)` is true |
+| `JMPFALSE` | iAsBx | Jump if `R(A)` is false |
+| `JMPNULL` | iAsBx | Jump if `R(A)` is null |
+
+### Function Calls
+
+| Opcode | Format | Description |
+|--------|--------|-------------|
+| `CALL` | iABC | Call `R(A)` with `B` args starting at `R(A+1)`, `C`=keep result |
+| `RETURN` | iA | Return `R(A)` |
+| `RETNIL` | — | Return null |
+| `CLOSURE` | iABx | Create closure from function pool entry `Bx` |
+
+### Object / Array
+
+| Opcode | Format | Description |
+|--------|--------|-------------|
+| `NEWOBJECT` | iA | `R(A) = {}` |
+| `NEWARRAY` | iABC | `R(A) = array(B)` |
+| `PUSH` | iABC | Push `R(B)` to array `R(A)` |
+
+## JSCodeRegister
+
+The compiled output for a function:
+
+```c
+struct JSCodeRegister {
+  uint16_t arity;           // argument count
+  uint16_t nr_slots;        // total register count
+  uint32_t cpool_count;     // constant pool size
+  JSValue *cpool;           // constant pool
+  uint32_t instr_count;     // instruction count
+  MachInstr32 *instructions; // 32-bit instruction array
+  uint32_t func_count;      // nested function count
+  JSCodeRegister **functions; // nested function table
+  JSValue name;             // function name
+  uint16_t disruption_pc;   // exception handler offset
+};
+```
+
+The constant pool holds all non-immediate values referenced by `LOADK` instructions: strings, large numbers, and other constants.

docs/spec/mcode.md

@@ -0,0 +1,90 @@
+---
+title: "Mcode IR"
+description: "JSON-based intermediate representation"
+---
+
+## Overview
+
+Mcode is a JSON-based intermediate representation that can be interpreted directly. It represents the same operations as the Mach register VM but uses string-based instruction dispatch rather than binary opcodes. Mcode is intended as an intermediate step toward native code compilation.
+
+## Pipeline
+
+```
+Source → Tokenize → Parse (AST) → Mcode (JSON) → Interpret
+                                                → Compile to Mach (planned)
+                                                → Compile to native (planned)
+```
+
+Mcode is produced by the `JS_Mcode` compiler pass, which emits a cJSON tree. The mcode interpreter walks this tree directly, dispatching on instruction name strings.
+
+## JSMCode Structure
+
+```c
+struct JSMCode {
+  uint16_t nr_args;        // argument count
+  uint16_t nr_slots;       // register count
+  cJSON **instrs;          // pre-flattened instruction array
+  uint32_t instr_count;    // number of instructions
+
+  struct {
+    const char *name;      // label name
+    uint32_t index;        // instruction index
+  } *labels;
+  uint32_t label_count;
+
+  struct JSMCode **functions; // nested functions
+  uint32_t func_count;
+
+  cJSON *json_root;        // keeps JSON alive
+  const char *name;        // function name
+  const char *filename;    // source file
+  uint16_t disruption_pc;  // exception handler offset
+};
+```
+
+## Instruction Format
+
+Each instruction is a JSON array. The first element is the instruction name (string), followed by operands:
+
+```json
+["LOADK", 0, 42]
+["ADD", 2, 0, 1]
+["JMPFALSE", 3, "else_label"]
+["CALL", 0, 2, 1]
+```
+
+The instruction set mirrors the Mach VM opcodes — same operations, same register semantics, but with string dispatch instead of numeric opcodes.
+
+## Labels
+
+Control flow uses named labels instead of numeric offsets:
+
+```json
+["LABEL", "loop_start"]
+["ADD", 1, 1, 2]
+["JMPFALSE", 3, "loop_end"]
+["JMP", "loop_start"]
+["LABEL", "loop_end"]
+```
+
+Labels are collected into a name-to-index map during loading, enabling O(1) jump resolution.
+
+## Differences from Mach
+
+| Property | Mcode | Mach |
+|----------|-------|------|
+| Instructions | cJSON arrays | 32-bit binary |
+| Dispatch | String comparison | Switch on opcode byte |
+| Constants | Inline in JSON | Separate constant pool |
+| Jump targets | Named labels | Numeric offsets |
+| Memory | Heap (cJSON nodes) | Off-heap (malloc) |
+
+## Purpose
+
+Mcode serves as an inspectable, debuggable intermediate format:
+
+- **Human-readable** — the JSON representation can be printed and examined
+- **Language-independent** — any tool that produces the correct JSON can target the ƿit runtime
+- **Compilation target** — the Mach compiler can consume mcode as input, and future native code generators can work from the same representation
+
+The cost of string-based dispatch makes mcode slower than the binary Mach VM, so it is primarily useful during development and as a compilation intermediate rather than for production execution.

docs/spec/objects.md

@@ -0,0 +1,141 @@
+---
+title: "Object Types"
+description: "Heap object header format and types"
+---
+
+## Object Header
+
+Every heap-allocated object begins with a 64-bit header word (`objhdr_t`):
+
+```
+[capacity: 56 bits][flags: 5 bits][type: 3 bits]
+```
+
+### Type Field (bits 0-2)
+
+| Value | Type | Description |
+|-------|------|-------------|
+| 0 | `OBJ_ARRAY` | Dynamic array of JSValues |
+| 1 | `OBJ_BLOB` | Binary data (bits) |
+| 2 | `OBJ_TEXT` | Unicode text string |
+| 3 | `OBJ_RECORD` | Key-value object with prototype chain |
+| 4 | `OBJ_FUNCTION` | Function (C, register, or mcode) |
+| 5 | `OBJ_CODE` | Compiled code |
+| 6 | `OBJ_FRAME` | Stack frame for closures |
+| 7 | `OBJ_FORWARD` | Forwarding pointer (GC) |
+
+### Flags (bits 3-7)
+
+- **Bit 3 (S)** — Stone flag. If set, the object is immutable and excluded from GC.
+- **Bit 4 (P)** — Properties flag.
+- **Bit 5 (A)** — Array flag.
+- **Bit 7 (R)** — Reserved.
+
+### Capacity (bits 8-63)
+
+The interpretation of the 56-bit capacity field depends on the object type.
+
+## Array
+
+```c
+struct JSArray {
+  objhdr_t header;     // type=0, capacity=element slots
+  word_t len;          // current number of elements
+  JSValue values[];    // inline flexible array
+};
+```
+
+Capacity is the number of JSValue slots allocated. Length is the number currently in use. Arrays grow by reallocating with a larger capacity.
+
+## Blob
+
+```c
+struct JSBlob {
+  objhdr_t header;     // type=1, capacity=allocated bits
+  word_t length;       // length in bits
+  uint8_t bits[];      // bit-packed data
+};
+```
+
+Blobs are bit-addressable. The length field tracks the exact number of bits written. A blob starts as antestone (mutable) for writing, then becomes stone (immutable) for reading.
+
+## Text
+
+```c
+struct JSText {
+  objhdr_t header;     // type=2, capacity=character slots
+  word_t length;       // length in codepoints (or hash if stoned)
+  word_t packed[];     // two UTF-32 chars per 64-bit word
+};
+```
+
+Text is stored as UTF-32, with two 32-bit codepoints packed per 64-bit word. When a text object is stoned, the length field is repurposed to cache the hash value (computed via `fash64`), since stoned text is immutable and the hash never changes.
+
+## Record
+
+```c
+struct JSRecord {
+  objhdr_t header;     // type=3, capacity=hash table slots
+  JSRecord *proto;     // prototype chain pointer
+  word_t len;          // number of entries
+  slot slots[];        // key-value pairs (hash table)
+};
+```
+
+Records use a hash table with linear probing. Slot 0 is reserved for internal metadata (class ID and record ID). Empty slots use `JS_NULL` as the key; deleted slots use `JS_EXCEPTION` as a tombstone.
+
+The prototype chain is a linked list of JSRecord pointers, traversed during property lookup.
+
+## Function
+
+```c
+struct JSFunction {
+  objhdr_t header;     // type=4
+  JSValue name;        // function name
+  int16_t length;      // arity (-1 for variadic)
+  uint8_t kind;        // C, register, or mcode
+  union {
+    struct { ... } cfunc;      // C function pointer
+    struct { ... } regvm;      // register VM code
+    struct { ... } mcode;      // mcode IR
+  } u;
+};
+```
+
+The kind field selects which union variant is active. Functions can be implemented in C (native), register code (mach VM), or mcode (JSON interpreter).
+
+## Frame
+
+```c
+struct JSFrame {
+  objhdr_t header;     // type=6, capacity=slot count
+  JSValue function;    // owning function
+  JSValue caller;      // parent frame
+  uint32_t return_pc;  // return address
+  JSValue slots[];     // [this][args][captured][locals][temps]
+};
+```
+
+Frames capture the execution context for closures. The slots array contains the function's `this` binding, arguments, captured upvalues, local variables, and temporaries. Frames are linked via the caller field for upvalue resolution across closure depth.
+
+## Forwarding Pointer
+
+```
+[pointer: 61 bits][111]
+```
+
+During garbage collection, when an object is copied to the new heap, the old header is replaced with a forwarding pointer to the new location. This is type 7 (`OBJ_FORWARD`) and stores the new address in bits 3-63. See [Garbage Collection](#gc) for details.
+
+## Object Sizing
+
+All objects are aligned to 8 bytes. The total size in bytes for each type:
+
+| Type | Size |
+|------|------|
+| Array | `8 + 8 + capacity * 8` |
+| Blob | `8 + 8 + ceil(capacity / 8)` |
+| Text | `8 + 8 + ceil(capacity / 2) * 8` |
+| Record | `8 + 8 + 8 + (capacity + 1) * 16` |
+| Function | `sizeof(JSFunction)` (fixed) |
+| Code | `sizeof(JSFunctionBytecode)` (fixed) |
+| Frame | `8 + 8 + 8 + 4 + capacity * 8` |

docs/spec/stone.md

@@ -0,0 +1,82 @@
+---
+title: "Stone Memory"
+description: "Immutable arena allocation"
+---
+
+## Overview
+
+Stone memory is a separate allocation arena for immutable values. Objects in stone memory are permanent — they are never moved, never freed, and never touched by the garbage collector.
+
+The `stone()` function in ƿit petrifies a value, deeply freezing it and all its descendants. Stoned objects have the S bit set in their object header.
+
+## The Stone Arena
+
+Stone memory uses bump allocation from a contiguous arena:
+
+```
+stone_base ──────── stone_free ──────── stone_end
+[allocated objects] [free space        ]
+```
+
+Allocation advances `stone_free` forward. When the arena is exhausted, overflow pages are allocated via the system allocator and linked together:
+
+```c
+struct StonePage {
+  struct StonePage *next;
+  size_t size;
+  uint8_t data[];
+};
+```
+
+## The S Bit
+
+Bit 3 of the object header is the stone flag. When set:
+
+- The object is **immutable** — writes disrupt
+- The object is **excluded from GC** — the collector skips it entirely
+- For text objects, the length field caches the **hash** instead of the character count (since the text cannot change, the hash is computed once and reused)
+
+## What Gets Stoned
+
+When `stone(value)` is called:
+
+1. If the value is already stone, return immediately
+2. Recursively walk all nested values (array elements, record fields, etc.)
+3. Copy each mutable object into the stone arena
+4. Set the S bit on each copied object
+5. Return the stoned value
+
+The operation is deep — an entire object graph becomes permanently immutable.
+
+## Text Interning
+
+The stone arena maintains a hash table for text interning. When a text value is stoned, it is looked up in the intern table. If an identical string already exists in stone memory, the existing one is reused. This deduplicates strings and makes equality comparison O(1) for stoned text.
+
+The hash is computed with `fash64` over the packed UTF-32 words.
+
+## Usage Patterns
+
+### Module Return Values
+
+Every module's return value is automatically stoned:
+
+```javascript
+// config.cm
+return {
+  debug: true,
+  timeout: 30
+}
+// The returned object is stone — shared safely between actors
+```
+
+### Message Passing
+
+Messages between actors are stoned before delivery, ensuring actors never share mutable state.
+
+### Constants
+
+Literal objects and arrays that can be determined at compile time may be allocated directly in stone memory.
+
+## Relationship to GC
+
+The Cheney copying collector only operates on the mutable heap. During collection, when the collector encounters a pointer to stone memory (S bit set), it skips it — stone objects are roots that never move. This means stone memory acts as a permanent root set with zero GC overhead.

docs/spec/values.md

@@ -0,0 +1,96 @@
+---
+title: "Value Representation"
+description: "JSValue tagging and encoding"
+---
+
+## Overview
+
+Every value in ƿit is a 64-bit word called a JSValue. The runtime uses LSB (least significant bit) tagging to pack type information directly into the value, avoiding heap allocation for common types.
+
+## Tag Encoding
+
+The lowest bits of a JSValue determine its type:
+
+| LSB Pattern | Type | Payload |
+|-------------|------|---------|
+| `xxxxxxx0` | Integer | 31-bit signed integer in upper bits |
+| `xxxxx001` | Pointer | 61-bit aligned heap pointer |
+| `xxxxx101` | Short float | 8-bit exponent + 52-bit mantissa |
+| `xxxxx011` | Special | 5-bit tag selects subtype |
+
+### Integers
+
+If the least significant bit is 0, the value is an immediate 31-bit signed integer. The integer is stored in the upper bits, extracted via `v >> 1`.
+
+```
+[integer: 31 bits][0]
+```
+
+Range: -1073741824 to 1073741823. Numbers outside this range are stored as short floats or heap-allocated.
+
+### Pointers
+
+If the lowest 3 bits are `001`, the value is a pointer to a heap object. The pointer is 8-byte aligned, so the low 3 bits are available for the tag. The actual address is extracted by clearing the low 3 bits.
+
+```
+[pointer: 61 bits][001]
+```
+
+All heap objects (arrays, records, blobs, text, functions, etc.) are referenced through pointer-tagged JSValues.
+
+### Short Floats
+
+If the lowest 3 bits are `101`, the value encodes a floating-point number directly. The format uses an 8-bit exponent (bias 127) and 52-bit mantissa, similar to IEEE 754 but with reduced range.
+
+```
+[sign: 1][exponent: 8][mantissa: 52][101]
+```
+
+Range: approximately ±3.4 * 10^38. Numbers outside this range fall back to null. Zero is always positive zero.
+
+### Specials
+
+If the lowest 2 bits are `11`, the next 3 bits select a special type:
+
+| 5-bit Tag | Value |
+|-----------|-------|
+| `00011` | Boolean (true/false in upper bits) |
+| `00111` | Null |
+| `01111` | Exception marker |
+| `10111` | Uninitialized |
+| `11011` | Immediate string |
+| `11111` | Catch offset |
+
+## Immediate Strings
+
+Short ASCII strings (up to 7 characters) are packed directly into the JSValue without heap allocation:
+
+```
+[char6][char5][char4][char3][char2][char1][char0][length: 3][11011]
+```
+
+Each character occupies 8 bits. The length (0-7) is stored in bits 5-7. Only ASCII characters (0-127) qualify — any non-ASCII character forces heap allocation.
+
+```javascript
+var s = "hello"   // 5 chars, fits in immediate string
+var t = ""         // immediate (length 0)
+var u = "longtext" // 8 chars, heap-allocated
+```
+
+## Null
+
+Null is encoded as a special-tagged value with tag `00111`. There is no `undefined` in ƿit — only null.
+
+```javascript
+var x = null       // special tag null
+var y = 1 / 0      // also null (division by zero)
+var z = {}.missing // null (missing field)
+```
+
+## Boolean
+
+True and false are encoded as specials with tag `00011`, distinguished by a bit in the upper payload.
+
+## Summary
+
+The tagging scheme ensures that the most common values — small integers, booleans, null, and short strings — require zero heap allocation. This significantly reduces GC pressure and improves cache locality.

docs/wota.md

@@ -0,0 +1,119 @@
+---
+title: "Wota Format"
+description: "Word Object Transfer Arrangement"
+weight: 86
+type: "docs"
+---
+
+Wota is a binary message format for local inter-process communication. It is similar to Nota but works at word granularity (64-bit words) rather than byte granularity. Wota arrangements are less compact than Nota but faster to arrange and consume.
+
+Wota stands for Word Object Transfer Arrangement.
+
+## Type Summary
+
+| Byte | Type |
+|------|------|
+| `00` | Integer |
+| `01` | Floating Point |
+| `02` | Array |
+| `03` | Record |
+| `04` | Blob |
+| `05` | Text |
+| `07` | Symbol |
+
+## Preambles
+
+Every Wota value starts with a preamble word. The least significant byte contains the type. The remaining 56 bits contain type-specific data.
+
+## Blob
+
+A blob is a string of bits. The remaining field contains the number of bits. The number of words that follow: `floor((number_of_bits + 63) / 64)`. The first bit of the blob goes into the most significant bit of the first word. The final word is padded with 0.
+
+Example: A blob containing 25 bits `111100001110001100100001`:
+
+```
+0000000000001904    # preamble: 25 bits, type blob
+F0E3208000000000    # data (padded to 64 bits)
+```
+
+## Text
+
+The text is a string of UTF-32 characters packed 2 per word. The remaining field contains the number of characters. The number of words that follow: `floor((number_of_characters + 1) / 2)`. The final word is padded with 0.
+
+Example: `"cat"`:
+
+```
+0000000000000305    # preamble: 3 characters, type text
+0000006300000061    # 'c' and 'a'
+0000007400000000    # 't' and padding
+```
+
+## Array
+
+An array is an ordered sequence of values. The remaining field contains the number of elements. Following the preamble are the elements, each beginning with its own preamble. Nesting is encouraged. Cyclic structures are not allowed.
+
+Example: `["duck", "dragon"]`:
+
+```
+0000000000000202    # preamble: 2 elements, type array
+0000000000000405    # text "duck": 4 chars
+0000006400000074    # 'd' 't' (reversed pair order)
+000000630000006B    # 'c' 'k'
+0000000000000605    # text "dragon": 6 chars
+0000006400000072    # 'd' 'r'
+0000006100000067    # 'a' 'g'
+0000006F0000006E    # 'o' 'n'
+```
+
+## Record
+
+A record is a set of key/value pairs. Keys must be text. The remaining field contains the number of pairs.
+
+Example: `{"ox": ["O", "X"]}`:
+
+```
+0000000000000103    # preamble: 1 pair, type record
+0000000000000205    # key "ox": 2 chars
+0000006F00000078    # 'o' 'x'
+0000000000000202    # value: array of 2
+0000000000000105    # "O": 1 char
+0000004F00000000    # 'O'
+0000000000000105    # "X": 1 char
+0000005800000000    # 'X'
+```
+
+## Number
+
+Numbers are represented as DEC64. To arrange an integer, shift the integer up 8 bits. The number is incorporated directly into the preamble.
+
+Example: `7`:
+
+```
+0000000000000700    # integer 7 as DEC64
+```
+
+To arrange a floating point number, place the number in the word following the floating point preamble.
+
+Example: `4.25`:
+
+```
+0000000000000001    # preamble: type floating point
+000000000001A9FE    # DEC64 encoding of 4.25
+```
+
+Care must be taken when decoding that the least significant byte of the number is not `80` (the null exponent).
+
+## Symbol
+
+The remaining field contains the symbol.
+
+Example: `[null, false, true, private, system]`:
+
+```
+0000000000000502    # array of 5
+0000000000000007    # null
+0000000000000207    # false
+0000000000000307    # true
+0000000000000807    # private
+0000000000000907    # system
+```

editors/ai/pit-context.md

@@ -0,0 +1,251 @@
+# ƿit Language — AI Context
+
+ƿit (pronounced "pit") is a safe, actor-based programming language. Its syntax resembles JavaScript but with significant differences. Scripts use `.ce` (actors) and `.cm` (modules) file extensions.
+
+## Key Differences from JavaScript
+
+- **`var` / `def`** — `var` is mutable, `def` is constant. No `let` or `const`.
+- **`==` is strict** — No `===` or `!==`. `==` and `!=` are always strict comparison.
+- **No `undefined`** — Only `null`. Division by zero produces `null`, not `Infinity`.
+- **No classes** — Use `meme()`, `proto()`, `isa()` for prototype chains.
+- **No `for...in`, `for...of`, spread, rest, or default params.**
+- **Variables declared at function body level only** — Not inside `if`/`while`/`for` blocks.
+- **All variables must be initialized** — `var x` alone is an error; use `var x = null`.
+- **`disrupt` / `disruption`** — No `try`/`catch`/`throw`. Error handling uses:
+  ```javascript
+  var fn = function() {
+    disrupt  // raise an error (bare keyword, no value)
+  } disruption {
+    // handle the error
+  }
+  ```
+- **No arraybuffers** — Use `blob` (works with bits; `stone(blob)` before reading).
+- **Identifiers can contain `?` and `!`** — e.g., `nil?`, `set!`, `is?valid`.
+- **4-parameter limit** — Functions take at most 4 named parameters.
+- **Everything lowercase** — Convention is all-lowercase identifiers with underscores.
+
+## Variable Declaration
+
+```javascript
+var count = 0         // mutable
+def MAX = 100         // constant (cannot be reassigned)
+var x = null          // must initialize (var x alone is an error)
+```
+
+## Functions
+
+```javascript
+var greet = function(name) {
+  print(`hello ${name}`)
+}
+
+// Arrow functions
+var double = x => x * 2
+var add = (a, b) => a + b
+```
+
+## Push / Pop Syntax
+
+```javascript
+var a = [1, 2]
+a[] = 3       // push: a is now [1, 2, 3]
+var v = a[]   // pop: v is 3, a is [1, 2]
+```
+
+## Control Flow
+
+```javascript
+if (x > 0) {
+  print("positive")
+} else {
+  print("non-positive")
+}
+
+while (i < 10) {
+  i = i + 1
+}
+
+for (var i = 0; i < 10; i = i + 1) {
+  print(i)
+}
+
+// do-while
+do {
+  i = i + 1
+} while (i < 10)
+```
+
+## Error Handling
+
+```javascript
+var safe_divide = function(a, b) {
+  if (b == 0) {
+    disrupt
+  }
+  return a / b
+} disruption {
+  return null
+}
+```
+
+## Creator Functions (Polymorphic)
+
+These examine argument types to decide behavior:
+
+### array()
+- `array(5)` — `[null, null, null, null, null]`
+- `array(3, 0)` — `[0, 0, 0]`
+- `array(5, i => i * 2)` — `[0, 2, 4, 6, 8]`
+- `array([1,2])` — copy
+- `array([1,2,3], x => x * 10)` — map: `[10, 20, 30]`
+- `array([1,2], [3,4])` — concat: `[1, 2, 3, 4]`
+- `array([1,2,3,4,5], 1, 4)` — slice: `[2, 3, 4]`
+- `array({a: 1, b: 2})` — keys: `["a", "b"]`
+- `array("hello")` — characters: `["h", "e", "l", "l", "o"]`
+- `array("a,b,c", ",")` — split: `["a", "b", "c"]`
+
+### text()
+- `text([1, 2, 3], ", ")` — join: `"1, 2, 3"`
+- `text(255, 16)` — radix: `"ff"`
+- `text("hello", 0, 3)` — substring: `"hel"`
+
+### number()
+- `number("42")` — parse: `42`
+- `number("ff", 16)` — radix: `255`
+- `number(true)` — `1`
+
+### record()
+- `record({a: 1})` — copy
+- `record({a: 1}, {b: 2})` — merge: `{a: 1, b: 2}`
+- `record(["x", "y"])` — from keys: `{x: true, y: true}`
+
+## All Intrinsic Functions
+
+**Constants:** `false`, `true`, `null`, `pi`
+
+**Type checks:** `is_array`, `is_blob`, `is_character`, `is_data`, `is_digit`, `is_false`, `is_fit`, `is_function`, `is_integer`, `is_letter`, `is_logical`, `is_lower`, `is_null`, `is_number`, `is_object`, `is_pattern`, `is_stone`, `is_text`, `is_true`, `is_upper`, `is_whitespace`
+
+**Creators:** `array`, `logical`, `number`, `record`, `text`
+
+**Math:** `abs`, `ceiling`, `floor`, `fraction`, `max`, `min`, `modulo`, `neg`, `remainder`, `round`, `sign`, `trunc`, `whole`
+
+**Text:** `character`, `codepoint`, `ends_with`, `extract`, `format`, `lower`, `normalize`, `replace`, `search`, `starts_with`, `trim`, `upper`
+
+**Array:** `every`, `filter`, `find`, `for`, `length`, `reduce`, `reverse`, `some`, `sort`
+
+**Objects:** `meme`, `proto`, `isa`, `stone`
+
+**Functions:** `apply`, `splat`
+
+**I/O:** `print`
+
+**Async:** `fallback`, `parallel`, `race`, `sequence`
+
+**Misc:** `logical`, `not`, `use`
+
+## Variable Scoping
+
+Variables are scoped to the function body in which they are declared. There is no block scoping. All declarations must be at the top level of a function body (not nested inside `if`/`while`/`for`).
+
+```javascript
+var outer = function() {
+  var x = 10
+  var inner = function() {
+    // x is visible here via closure
+    print(x)
+  }
+  inner()
+}
+```
+
+## Modules (.cm files)
+
+Modules return a value (typically a record of exports). They are loaded with `use()`, cached, and frozen.
+
+```javascript
+// math_utils.cm
+var square = x => x * x
+var cube = x => x * x * x
+return {square: square, cube: cube}
+
+// main.ce
+var utils = use('math_utils')
+print(utils.square(5))  // 25
+```
+
+## Standard Library (loaded with use())
+
+- `blob` — binary data (works with bits, not bytes)
+- `time` — time constants and conversions
+- `math` — trig, logarithms, roots (sub-modules: `math/radians`, `math/turns`)
+- `json` — JSON encoding/decoding (`json.encode`, `json.decode`)
+- `random` — random number generation
+- `fd` — file descriptor operations (`fd.read`, `fd.write`, `fd.slurp`, `fd.stat`)
+
+## Actor Model (.ce files)
+
+Actors are independent execution units that never share memory. They communicate via message passing.
+
+```javascript
+// greeter.ce
+$receiver(function(msg) {
+  $send(msg.from, {greeting: `hello ${msg.name}`})
+})
+```
+
+### Actor Intrinsics ($ prefix)
+
+- `$me` — this actor's address
+- `$send(address, message)` — send a message
+- `$start(script, env)` — start a new actor
+- `$stop()` — stop this actor
+- `$delay(ms)` — delay processing
+- `$receiver(fn)` — set message handler
+- `$clock(interval, message)` — periodic self-message
+- `$portal(name)` — create named portal
+- `$contact(name)` — connect to portal
+- `$couple(address)` — lifecycle coupling
+- `$unneeded(fn)` — cleanup callback
+- `$connection(address)` — establish connection
+- `$time_limit(ms)` — execution time limit
+
+## Common Patterns
+
+### Iteration
+```javascript
+// Preferred: use for() intrinsic
+for([1, 2, 3], function(item, index) {
+  print(`${text(index)}: ${text(item)}`)
+})
+
+// C-style for loop
+for (var i = 0; i < length(items); i = i + 1) {
+  print(items[i])
+}
+```
+
+### String Building
+```javascript
+// Use backtick interpolation
+var msg = `hello ${name}, you are ${text(age)} years old`
+
+// Join array
+var csv = text(values, ",")
+```
+
+### Record Manipulation
+```javascript
+var obj = {name: "alice", age: 30}
+var keys = array(obj)           // ["name", "age"]
+var copy = record(obj)          // mutable copy
+var merged = record(obj, {role: "admin"})
+```
+
+### Error-Safe Operations
+```javascript
+var safe_parse = function(input) {
+  return number(input)
+} disruption {
+  return null
+}
+```

editors/vscode/language-configuration.json

@@ -0,0 +1,30 @@
+{
+  "comments": {
+    "lineComment": "//",
+    "blockComment": ["/*", "*/"]
+  },
+  "brackets": [
+    ["{", "}"],
+    ["[", "]"],
+    ["(", ")"]
+  ],
+  "autoClosingPairs": [
+    { "open": "{", "close": "}" },
+    { "open": "[", "close": "]" },
+    { "open": "(", "close": ")" },
+    { "open": "\"", "close": "\"", "notIn": ["string"] },
+    { "open": "`", "close": "`", "notIn": ["string"] }
+  ],
+  "surroundingPairs": [
+    ["{", "}"],
+    ["[", "]"],
+    ["(", ")"],
+    ["\"", "\""],
+    ["`", "`"]
+  ],
+  "indentationRules": {
+    "increaseIndentPattern": "^.*\\{[^}\"'`]*$",
+    "decreaseIndentPattern": "^\\s*\\}"
+  },
+  "wordPattern": "[a-zA-Z_$][a-zA-Z0-9_$?!]*"
+}

editors/vscode/lsp/analysis.cm

@@ -0,0 +1,113 @@
+// Document analysis module.
+// Call make(tokenize_mod, parse_mod) to get an analysis object.
+
+var json = use('json')
+
+// Create an analysis module bound to the tokenize and parse functions.
+var make = function(tokenize_mod, parse_mod) {
+
+  // Tokenize and parse a document, storing the results.
+  var update = function(docs, uri, params) {
+    var src = params.src
+    var version = params.version
+    var tok_result = null
+    var ast = null
+    var errors = []
+    var doc = null
+
+    var do_tokenize = function() {
+      tok_result = tokenize_mod(src, uri)
+    } disruption {
+      errors = [{message: "Tokenize failed", line: 1, column: 1}]
+    }
+    var do_parse = function() {
+      ast = parse_mod(tok_result.tokens, src, uri, tokenize_mod)
+    } disruption {
+      // parse_mod may set errors on ast even on partial failure
+    }
+
+    do_tokenize()
+
+    if (tok_result != null) {
+      do_parse()
+
+      if (ast != null && ast.errors != null) {
+        errors = ast.errors
+      }
+    }
+
+    doc = {
+      uri: uri,
+      text: src,
+      version: version,
+      tokens: (tok_result != null) ? tok_result.tokens : [],
+      ast: ast,
+      errors: errors
+    }
+    docs[uri] = doc
+    return doc
+  }
+
+  // Remove a document from the store.
+  var remove = function(docs, uri) {
+    delete docs[uri]
+  }
+
+  // Convert parse errors to LSP diagnostics.
+  var diagnostics = function(doc) {
+    var result = []
+    var _i = 0
+    var e = null
+    var line = null
+    var col = null
+    while (_i < length(doc.errors)) {
+      e = doc.errors[_i]
+      line = (e.line != null) ? e.line - 1 : 0
+      col = (e.column != null) ? e.column - 1 : 0
+      result[] = {
+        range: {
+          start: {line: line, character: col},
+          end: {line: line, character: col + 1}
+        },
+        severity: 1,
+        source: "pit",
+        message: e.message
+      }
+      _i = _i + 1
+    }
+    return result
+  }
+
+  // Find the token at a given line/column (0-based).
+  var token_at = function(doc, line, col) {
+    var tokens = doc.tokens
+    var _i = 0
+    var tok = null
+    while (_i < length(tokens)) {
+      tok = tokens[_i]
+      if (tok.from_row == line && tok.from_column <= col && tok.to_column >= col) {
+        return tok
+      }
+      if (tok.from_row < line && tok.to_row > line) {
+        return tok
+      }
+      if (tok.from_row < line && tok.to_row == line && tok.to_column >= col) {
+        return tok
+      }
+      if (tok.from_row == line && tok.to_row > line && tok.from_column <= col) {
+        return tok
+      }
+      _i = _i + 1
+    }
+    return null
+  }
+
+  return {
+    update: update,
+    remove: remove,
+    diagnostics: diagnostics,
+    token_at: token_at
+  }
+}
+
+return make

editors/vscode/lsp/completions.cm

@@ -0,0 +1,133 @@
+// Completion provider for the ƿit LSP.
+
+// CompletionItemKind constants (LSP spec)
+def KIND_FUNCTION = 3
+def KIND_VARIABLE = 6
+def KIND_KEYWORD = 14
+def KIND_CONSTANT = 21
+
+// All intrinsic function names
+def intrinsic_functions = [
+  "abs", "apply", "array", "ceiling", "character", "codepoint",
+  "ends_with", "every", "extract", "fallback", "filter", "find",
+  "floor", "format", "fraction",
+  "is_array", "is_blob", "is_character", "is_data", "is_digit",
+  "is_false", "is_fit", "is_function", "is_integer", "is_letter",
+  "is_logical", "is_lower", "is_null", "is_number", "is_object",
+  "is_pattern", "is_stone", "is_text", "is_true", "is_upper",
+  "is_whitespace",
+  "length", "logical", "lower", "max", "min", "modulo",
+  "neg", "normalize", "not", "number",
+  "parallel", "print", "race", "record", "reduce", "remainder",
+  "replace", "reverse", "round",
+  "search", "sequence", "sign", "some", "sort", "starts_with",
+  "stone", "text", "trim", "trunc", "upper", "whole",
+  "meme", "proto", "isa", "splat", "use"
+]
+
+// Keywords that can be completed
+def keywords = [
+  "var", "def", "if", "else", "for", "while", "do",
+  "function", "return", "go", "break", "continue",
+  "disrupt", "disruption", "delete", "in", "this",
+  "null", "true", "false"
+]
+
+// Actor intrinsics (only in .ce files)
+def actor_intrinsics = [
+  "$me", "$send", "$start", "$stop", "$delay",
+  "$receiver", "$clock", "$portal", "$contact",
+  "$couple", "$unneeded", "$connection", "$time_limit"
+]
+
+// Walk AST scopes to find variables visible at a position.
+var collect_scope_vars = function(doc, line, col) {
+  var vars = []
+  var ast = doc.ast
+  var _i = 0
+  var _j = 0
+  var scope = null
+  var v = null
+
+  if (ast == null || ast.scopes == null) {
+    return vars
+  }
+
+  // Collect variables from all scopes (simplified: return all declared vars)
+  while (_i < length(ast.scopes)) {
+    scope = ast.scopes[_i]
+    if (scope.vars != null) {
+      _j = 0
+      while (_j < length(scope.vars)) {
+        v = scope.vars[_j]
+        if (v.name != null) {
+          vars[] = {
+            label: v.name,
+            kind: (v.is_const == true) ? KIND_CONSTANT : KIND_VARIABLE,
+            detail: (v.is_const == true) ? "def" : "var"
+          }
+        }
+        _j = _j + 1
+      }
+    }
+    _i = _i + 1
+  }
+
+  return vars
+}
+
+// Provide completions for a document at a position.
+var complete = function(doc, line, col) {
+  var items = []
+  var _i = 0
+  var is_actor = ends_with(doc.uri, ".ce")
+
+  // Intrinsic functions
+  _i = 0
+  while (_i < length(intrinsic_functions)) {
+    items[] = {
+      label: intrinsic_functions[_i],
+      kind: KIND_FUNCTION,
+      detail: "intrinsic"
+    }
+    _i = _i + 1
+  }
+
+  // Keywords
+  _i = 0
+  while (_i < length(keywords)) {
+    items[] = {
+      label: keywords[_i],
+      kind: KIND_KEYWORD,
+      detail: "keyword"
+    }
+    _i = _i + 1
+  }
+
+  // Actor intrinsics (only for .ce files)
+  if (is_actor) {
+    _i = 0
+    while (_i < length(actor_intrinsics)) {
+      items[] = {
+        label: actor_intrinsics[_i],
+        kind: KIND_FUNCTION,
+        detail: "actor intrinsic"
+      }
+      _i = _i + 1
+    }
+  }
+
+  // Variables from scope analysis
+  var scope_vars = collect_scope_vars(doc, line, col)
+  _i = 0
+  while (_i < length(scope_vars)) {
+    items[] = scope_vars[_i]
+    _i = _i + 1
+  }
+
+  return items
+}
+
+return {
+  complete: complete
+}

editors/vscode/lsp/hover.cm

@@ -0,0 +1,461 @@
+// Hover provider for the ƿit LSP.
+// Shows documentation for intrinsic functions and variable info.
+
+// Intrinsic function documentation database.
+// Each entry: {signature, description}
+def intrinsic_docs = {
+  abs: {
+    signature: "abs(number)",
+    description: "Absolute value. Returns null for non-numbers."
+  },
+  apply: {
+    signature: "apply(function, array)",
+    description: "Execute the function, passing array elements as input values."
+  },
+  array: {
+    signature: "array(value, ...)",
+    description: "Create arrays. Polymorphic: array(number) creates sized array, array(array) copies, array(array, fn) maps, array(text) splits into characters, array(text, sep) splits by separator."
+  },
+  ceiling: {
+    signature: "ceiling(number, place)",
+    description: "Round up. If place is 0 or null, round to smallest integer >= number."
+  },
+  character: {
+    signature: "character(value)",
+    description: "If text, returns the first character. If a non-negative integer, returns the character from that codepoint."
+  },
+  codepoint: {
+    signature: "codepoint(text)",
+    description: "Returns the codepoint number of the first character."
+  },
+  ends_with: {
+    signature: "ends_with(text, suffix)",
+    description: "Returns true if the text ends with the given suffix."
+  },
+  every: {
+    signature: "every(array, function)",
+    description: "Returns true if every element satisfies the predicate."
+  },
+  extract: {
+    signature: "extract(text, pattern, from, to)",
+    description: "Match text to pattern. Returns a record of saved fields, or null if no match."
+  },
+  fallback: {
+    signature: "fallback(requestor_array)",
+    description: "Returns a requestor that tries each requestor in order until one succeeds."
+  },
+  filter: {
+    signature: "filter(array, function)",
+    description: "Returns a new array containing elements for which function returns true."
+  },
+  find: {
+    signature: "find(array, function, reverse, from)",
+    description: "Returns the element number where function returns true, or null if not found. If second arg is not a function, compares directly."
+  },
+  floor: {
+    signature: "floor(number, place)",
+    description: "Round down. If place is 0 or null, round to greatest integer <= number."
+  },
+  format: {
+    signature: "format(text, collection, transformer)",
+    description: "Substitute {key} placeholders in text with values from a collection (array or record)."
+  },
+  fraction: {
+    signature: "fraction(number)",
+    description: "Returns the fractional part of a number."
+  },
+  is_array: {
+    signature: "is_array(value)",
+    description: "Returns true if the value is an array."
+  },
+  is_blob: {
+    signature: "is_blob(value)",
+    description: "Returns true if the value is a blob."
+  },
+  is_character: {
+    signature: "is_character(value)",
+    description: "Returns true if the value is a single character."
+  },
+  is_data: {
+    signature: "is_data(value)",
+    description: "Returns true if the value is data (not a function)."
+  },
+  is_digit: {
+    signature: "is_digit(value)",
+    description: "Returns true if the value is a digit character."
+  },
+  is_false: {
+    signature: "is_false(value)",
+    description: "Returns true if the value is false."
+  },
+  is_fit: {
+    signature: "is_fit(value)",
+    description: "Returns true if the value is a fit integer."
+  },
+  is_function: {
+    signature: "is_function(value)",
+    description: "Returns true if the value is a function."
+  },
+  is_integer: {
+    signature: "is_integer(value)",
+    description: "Returns true if the value is an integer."
+  },
+  is_letter: {
+    signature: "is_letter(value)",
+    description: "Returns true if the value is a letter character."
+  },
+  is_logical: {
+    signature: "is_logical(value)",
+    description: "Returns true if the value is a logical (boolean)."
+  },
+  is_lower: {
+    signature: "is_lower(value)",
+    description: "Returns true if the value is a lowercase character."
+  },
+  is_null: {
+    signature: "is_null(value)",
+    description: "Returns true if the value is null."
+  },
+  is_number: {
+    signature: "is_number(value)",
+    description: "Returns true if the value is a number."
+  },
+  is_object: {
+    signature: "is_object(value)",
+    description: "Returns true if the value is an object (record)."
+  },
+  is_pattern: {
+    signature: "is_pattern(value)",
+    description: "Returns true if the value is a pattern (regex)."
+  },
+  is_stone: {
+    signature: "is_stone(value)",
+    description: "Returns true if the value is frozen (stoned)."
+  },
+  is_text: {
+    signature: "is_text(value)",
+    description: "Returns true if the value is text."
+  },
+  is_true: {
+    signature: "is_true(value)",
+    description: "Returns true if the value is true."
+  },
+  is_upper: {
+    signature: "is_upper(value)",
+    description: "Returns true if the value is an uppercase character."
+  },
+  is_whitespace: {
+    signature: "is_whitespace(value)",
+    description: "Returns true if the value is a whitespace character."
+  },
+  length: {
+    signature: "length(value)",
+    description: "Array: number of elements. Text: number of codepoints. Function: arity. Blob: number of bits. Record: record.length()."
+  },
+  logical: {
+    signature: "logical(value)",
+    description: "Convert to logical. 0/false/null/\"false\" produce false; 1/true/\"true\" produce true."
+  },
+  lower: {
+    signature: "lower(text)",
+    description: "Returns text with all uppercase characters converted to lowercase."
+  },
+  max: {
+    signature: "max(number, number)",
+    description: "Returns the larger of two numbers."
+  },
+  min: {
+    signature: "min(number, number)",
+    description: "Returns the smaller of two numbers."
+  },
+  modulo: {
+    signature: "modulo(dividend, divisor)",
+    description: "Result has the sign of the divisor."
+  },
+  neg: {
+    signature: "neg(number)",
+    description: "Negate. Reverse the sign of a number."
+  },
+  normalize: {
+    signature: "normalize(text)",
+    description: "Unicode normalize."
+  },
+  not: {
+    signature: "not(logical)",
+    description: "Returns the opposite logical. Returns null for non-logicals."
+  },
+  number: {
+    signature: "number(value, radix_or_format)",
+    description: "Convert to number. Polymorphic: number(logical), number(text), number(text, radix), number(text, format)."
+  },
+  parallel: {
+    signature: "parallel(requestor_array, throttle, need)",
+    description: "Start all requestors concurrently. Optional throttle limits concurrency; optional need specifies minimum successes."
+  },
+  print: {
+    signature: "print(value)",
+    description: "Print a value to standard output."
+  },
+  race: {
+    signature: "race(requestor_array, throttle, need)",
+    description: "Like parallel but returns as soon as needed results are obtained. Default need is 1."
+  },
+  record: {
+    signature: "record(value, ...)",
+    description: "Create records. Polymorphic: record(record) copies, record(record, record) merges, record(array) creates from keys."
+  },
+  reduce: {
+    signature: "reduce(array, function, initial, reverse)",
+    description: "Reduce an array to a single value by applying a function to pairs of elements."
+  },
+  remainder: {
+    signature: "remainder(dividend, divisor)",
+    description: "For fit integers: dividend - ((dividend // divisor) * divisor)."
+  },
+  replace: {
+    signature: "replace(text, target, replacement, limit)",
+    description: "Return text with target replaced. Target can be text or pattern. Replacement can be text or function."
+  },
+  reverse: {
+    signature: "reverse(array)",
+    description: "Returns a new array with elements in the opposite order."
+  },
+  round: {
+    signature: "round(number, place)",
+    description: "Round to nearest."
+  },
+  search: {
+    signature: "search(text, target, from)",
+    description: "Search text for target. Returns character position or null."
+  },
+  sequence: {
+    signature: "sequence(requestor_array)",
+    description: "Process requestors in order. Each result becomes input to the next."
+  },
+  sign: {
+    signature: "sign(number)",
+    description: "Returns -1, 0, or 1."
+  },
+  some: {
+    signature: "some(array, function)",
+    description: "Returns true if any element satisfies the predicate."
+  },
+  sort: {
+    signature: "sort(array, select)",
+    description: "Returns a new sorted array. Sort keys must be all numbers or all texts. Ascending and stable."
+  },
+  starts_with: {
+    signature: "starts_with(text, prefix)",
+    description: "Returns true if the text starts with the given prefix."
+  },
+  stone: {
+    signature: "stone(value)",
+    description: "Petrify the value, making it permanently immutable. Deep freeze."
+  },
+  text: {
+    signature: "text(value, ...)",
+    description: "Convert to text. Polymorphic: text(array, sep) joins, text(number, radix/format) formats, text(text, from, to) substrings."
+  },
+  trim: {
+    signature: "trim(text, reject)",
+    description: "Remove characters from both ends. Default removes whitespace."
+  },
+  trunc: {
+    signature: "trunc(number, place)",
+    description: "Truncate toward zero."
+  },
+  upper: {
+    signature: "upper(text)",
+    description: "Returns text with all lowercase characters converted to uppercase."
+  },
+  whole: {
+    signature: "whole(number)",
+    description: "Returns the whole part of a number."
+  },
+  meme: {
+    signature: "meme()",
+    description: "Create a new meme (prototype chain marker)."
+  },
+  proto: {
+    signature: "proto(object, meme)",
+    description: "Set the prototype meme of an object."
+  },
+  isa: {
+    signature: "isa(object, meme)",
+    description: "Returns true if the object has the given meme in its prototype chain."
+  },
+  splat: {
+    signature: "splat(function, array)",
+    description: "Call function with array elements as separate arguments."
+  },
+  use: {
+    signature: "use(path)",
+    description: "Load a module. Returns the module's exported value. Modules are cached and frozen."
+  },
+  pi: {
+    signature: "pi",
+    description: "An approximation of circumference / diameter: 3.1415926535897932."
+  }
+}
+
+// Actor intrinsic documentation
+def actor_docs = {
+  "$me": {
+    signature: "$me",
+    description: "The address of this actor."
+  },
+  "$send": {
+    signature: "$send(address, message)",
+    description: "Send a message to another actor."
+  },
+  "$start": {
+    signature: "$start(script, env)",
+    description: "Start a new actor from a script path."
+  },
+  "$stop": {
+    signature: "$stop()",
+    description: "Stop this actor."
+  },
+  "$delay": {
+    signature: "$delay(milliseconds)",
+    description: "Delay processing for a number of milliseconds."
+  },
+  "$receiver": {
+    signature: "$receiver(function)",
+    description: "Set the message receiver function for this actor."
+  },
+  "$clock": {
+    signature: "$clock(interval, message)",
+    description: "Send a message to self at regular intervals."
+  },
+  "$portal": {
+    signature: "$portal(name)",
+    description: "Create a named portal for inter-actor communication."
+  },
+  "$contact": {
+    signature: "$contact(portal_name)",
+    description: "Connect to a named portal."
+  },
+  "$couple": {
+    signature: "$couple(address)",
+    description: "Couple with another actor for lifecycle management."
+  },
+  "$unneeded": {
+    signature: "$unneeded(function)",
+    description: "Set a function to be called when this actor is no longer needed."
+  },
+  "$connection": {
+    signature: "$connection(address)",
+    description: "Establish a connection with another actor."
+  },
+  "$time_limit": {
+    signature: "$time_limit(milliseconds)",
+    description: "Set a time limit for this actor's execution."
+  }
+}
+
+// Provide hover info for a token.
+var hover = function(doc, line, col, token_at) {
+  var tok = token_at(doc, line, col)
+  var info = null
+  var name = null
+  var _i = 0
+  var _j = 0
+  var scope = null
+  var v = null
+
+  if (tok == null) {
+    return null
+  }
+
+  // Check intrinsic functions
+  if (tok.kind == "name" && tok.value != null) {
+    name = tok.value
+    info = intrinsic_docs[name]
+    if (info != null) {
+      return {
+        contents: {
+          kind: "markdown",
+          value: `**${info.signature}**\n\n${info.description}`
+        }
+      }
+    }
+  }
+
+  // Check actor intrinsics ($name)
+  if (tok.value != null && starts_with(tok.value, "$")) {
+    info = actor_docs[tok.value]
+    if (info != null) {
+      return {
+        contents: {
+          kind: "markdown",
+          value: `**${info.signature}**\n\n${info.description}`
+        }
+      }
+    }
+  }
+
+  // Check keywords
+  if (tok.kind == "var" || tok.kind == "def") {
+    return {
+      contents: {
+        kind: "markdown",
+        value: (tok.kind == "var")
+          ? "**var** — Declare a mutable variable."
+          : "**def** — Declare a constant."
+      }
+    }
+  }
+
+  if (tok.kind == "disrupt") {
+    return {
+      contents: {
+        kind: "markdown",
+        value: "**disrupt** — Raise an error. Use with **disruption** block to handle errors."
+      }
+    }
+  }
+
+  if (tok.kind == "disruption") {
+    return {
+      contents: {
+        kind: "markdown",
+        value: "**disruption** — Error handling block. Catches errors raised by **disrupt**."
+      }
+    }
+  }
+
+  // User variable: show declaration info from scope
+  if (tok.kind == "name" && tok.value != null && doc.ast != null && doc.ast.scopes != null) {
+    _i = 0
+    while (_i < length(doc.ast.scopes)) {
+      scope = doc.ast.scopes[_i]
+      if (scope.vars != null) {
+        _j = 0
+        while (_j < length(scope.vars)) {
+          v = scope.vars[_j]
+          if (v.name == tok.value) {
+            return {
+              contents: {
+                kind: "markdown",
+                value: (v.is_const == true)
+                  ? `**def** ${v.name}`
+                  : `**var** ${v.name}`
+              }
+            }
+          }
+          _j = _j + 1
+        }
+      }
+      _i = _i + 1
+    }
+  }
+
+  return null
+}
+
+return {
+  hover: hover,
+  intrinsic_docs: intrinsic_docs,
+  actor_docs: actor_docs
+}

editors/vscode/lsp/lsp.ce

@@ -0,0 +1,209 @@
+// ƿit Language Server Protocol (LSP) main loop.
+// Communicates via JSON-RPC over stdin/stdout.
+
+var fd = use('fd')
+var json_mod = use('json')
+var protocol = use('protocol')
+var analysis_make = use('analysis')
+var completions = use('completions')
+var hover_mod = use('hover')
+var symbols = use('symbols')
+
+// Get tokenize_mod and parse_mod from the environment.
+// These are the same functions the compiler uses internally.
+var tokenize_mod = use('tokenize')
+var parse_mod = use('parse')
+
+// Create analysis module bound to tokenize/parse
+var analysis = analysis_make(tokenize_mod, parse_mod)
+
+// Document store: URI -> {text, version, ast, tokens, errors}
+var docs = {}
+
+// Log to stderr for debugging (does not interfere with protocol).
+var log = function(msg) {
+  fd.write(2, `[pit-lsp] ${msg}\n`)
+}
+
+// Publish diagnostics for a document.
+var publish_diagnostics = function(uri, doc) {
+  var diags = analysis.diagnostics(doc)
+  protocol.notify("textDocument/publishDiagnostics", {
+    uri: uri,
+    diagnostics: diags
+  })
+}
+
+// Parse a document and publish diagnostics.
+var parse_and_notify = function(uri, src, version) {
+  var doc = analysis.update(docs, uri, {src: src, version: version})
+  publish_diagnostics(uri, doc)
+}
+
+// Handle initialize request.
+var handle_initialize = function(id, params) {
+  protocol.respond(id, {
+    capabilities: {
+      textDocumentSync: {
+        openClose: true,
+        change: 1,
+        save: {includeText: true}
+      },
+      completionProvider: {
+        triggerCharacters: [".", "$"]
+      },
+      hoverProvider: true,
+      definitionProvider: true,
+      documentSymbolProvider: true
+    },
+    serverInfo: {
+      name: "pit-lsp",
+      version: "0.1.0"
+    }
+  })
+}
+
+// Handle textDocument/didOpen notification.
+var handle_did_open = function(params) {
+  var td = params.textDocument
+  parse_and_notify(td.uri, td.text, td.version)
+}
+
+// Handle textDocument/didChange notification (full text sync).
+var handle_did_change = function(params) {
+  var td = params.textDocument
+  var changes = params.contentChanges
+  if (length(changes) > 0) {
+    parse_and_notify(td.uri, changes[0].text, td.version)
+  }
+}
+
+// Handle textDocument/didClose notification.
+var handle_did_close = function(params) {
+  var uri = params.textDocument.uri
+  analysis.remove(docs, uri)
+  // Clear diagnostics
+  protocol.notify("textDocument/publishDiagnostics", {
+    uri: uri,
+    diagnostics: []
+  })
+}
+
+// Handle textDocument/didSave notification.
+var handle_did_save = function(params) {
+  var td = params.textDocument
+  if (params.text != null) {
+    parse_and_notify(td.uri, params.text, td.version)
+  }
+}
+
+// Handle textDocument/completion request.
+var handle_completion = function(id, params) {
+  var uri = params.textDocument.uri
+  var pos = params.position
+  var doc = docs[uri]
+  var items = []
+  if (doc != null) {
+    items = completions.complete(doc, pos.line, pos.character)
+  }
+  protocol.respond(id, items)
+}
+
+// Handle textDocument/hover request.
+var handle_hover = function(id, params) {
+  var uri = params.textDocument.uri
+  var pos = params.position
+  var doc = docs[uri]
+  var result = null
+  if (doc != null) {
+    result = hover_mod.hover(doc, pos.line, pos.character, analysis.token_at)
+  }
+  protocol.respond(id, result)
+}
+
+// Handle textDocument/definition request.
+var handle_definition = function(id, params) {
+  var uri = params.textDocument.uri
+  var pos = params.position
+  var doc = docs[uri]
+  var result = null
+  if (doc != null) {
+    result = symbols.definition(doc, pos.line, pos.character, analysis.token_at)
+  }
+  protocol.respond(id, result)
+}
+
+// Handle textDocument/documentSymbol request.
+var handle_document_symbol = function(id, params) {
+  var uri = params.textDocument.uri
+  var doc = docs[uri]
+  var result = []
+  if (doc != null) {
+    result = symbols.document_symbols(doc)
+  }
+  protocol.respond(id, result)
+}
+
+// Dispatch a single message. Wrapped in a function for disruption handling.
+var dispatch_message = function(msg) {
+  var method = msg.method
+  if (method == "initialize") {
+    handle_initialize(msg.id, msg.params)
+  } else if (method == "initialized") {
+    // no-op
+  } else if (method == "textDocument/didOpen") {
+    handle_did_open(msg.params)
+  } else if (method == "textDocument/didChange") {
+    handle_did_change(msg.params)
+  } else if (method == "textDocument/didClose") {
+    handle_did_close(msg.params)
+  } else if (method == "textDocument/didSave") {
+    handle_did_save(msg.params)
+  } else if (method == "textDocument/completion") {
+    handle_completion(msg.id, msg.params)
+  } else if (method == "textDocument/hover") {
+    handle_hover(msg.id, msg.params)
+  } else if (method == "textDocument/definition") {
+    handle_definition(msg.id, msg.params)
+  } else if (method == "textDocument/documentSymbol") {
+    handle_document_symbol(msg.id, msg.params)
+  } else if (method == "shutdown") {
+    protocol.respond(msg.id, null)
+    return "shutdown"
+  } else if (method == "exit") {
+    return "exit"
+  } else {
+    if (msg.id != null) {
+      protocol.respond_error(msg.id, -32601, `Method not found: ${method}`)
+    }
+  }
+  return null
+} disruption {
+  log(`error handling ${msg.method}`)
+  if (msg.id != null) {
+    protocol.respond_error(msg.id, -32603, `Internal error handling ${msg.method}`)
+  }
+  return null
+}
+
+// Main loop.
+log("starting")
+
+var running = true
+var msg = null
+var result = null
+
+while (running) {
+  msg = protocol.read_message()
+  if (msg == null) {
+    running = false
+    break
+  }
+
+  result = dispatch_message(msg)
+  if (result == "exit") {
+    running = false
+  }
+}
+
+log("stopped")

editors/vscode/lsp/protocol.cm

@@ -0,0 +1,102 @@
+// JSON-RPC protocol helpers for LSP communication over stdin/stdout.
+// Reads Content-Length framed messages from stdin, writes to stdout.
+
+var fd = use('fd')
+var json = use('json')
+
+// Read a single JSON-RPC message from stdin.
+// Protocol: "Content-Length: N\r\n\r\n" followed by N bytes of JSON.
+var read_message = function() {
+  var header = ""
+  var ch = null
+  var content_length = null
+  var body = null
+  var total = 0
+  var chunk = null
+
+  // Read header byte by byte until we hit \r\n\r\n
+  while (true) {
+    ch = fd.read(0, 1)
+    if (ch == null) {
+      return null
+    }
+    header = header + text(ch)
+    if (ends_with(header, "\r\n\r\n")) {
+      break
+    }
+  }
+
+  // Parse Content-Length from header
+  var lines = array(header, "\r\n")
+  var _i = 0
+  while (_i < length(lines)) {
+    if (starts_with(lines[_i], "Content-Length:")) {
+      content_length = number(trim(text(lines[_i], 16)))
+    }
+    _i = _i + 1
+  }
+
+  if (content_length == null) {
+    return null
+  }
+
+  // Read exactly content_length bytes
+  body = ""
+  total = 0
+  while (total < content_length) {
+    chunk = fd.read(0, content_length - total)
+    if (chunk == null) {
+      return null
+    }
+    chunk = text(chunk)
+    body = body + chunk
+    total = total + length(chunk)
+  }
+
+  return json.decode(body)
+}
+
+// Send a JSON-RPC message to stdout.
+var send_message = function(msg) {
+  var body = json.encode(msg)
+  var header = `Content-Length: ${text(length(body))}\r\n\r\n`
+  fd.write(1, header + body)
+}
+
+// Send a JSON-RPC response for a request.
+var respond = function(id, result) {
+  send_message({
+    jsonrpc: "2.0",
+    id: id,
+    result: result
+  })
+}
+
+// Send a JSON-RPC error response.
+var respond_error = function(id, code, message) {
+  send_message({
+    jsonrpc: "2.0",
+    id: id,
+    error: {
+      code: code,
+      message: message
+    }
+  })
+}
+
+// Send a JSON-RPC notification (no id).
+var notify = function(method, params) {
+  send_message({
+    jsonrpc: "2.0",
+    method: method,
+    params: params
+  })
+}
+
+return {
+  read_message: read_message,
+  send_message: send_message,
+  respond: respond,
+  respond_error: respond_error,
+  notify: notify
+}

editors/vscode/lsp/symbols.cm

@@ -0,0 +1,238 @@
+// Document symbols and go-to-definition provider for the ƿit LSP.
+
+// SymbolKind constants (LSP spec)
+def KIND_FUNCTION = 12
+def KIND_VARIABLE = 13
+def KIND_CONSTANT = 14
+
+// Walk AST to extract document symbols (top-level vars/defs and functions).
+var document_symbols = function(doc) {
+  var symbols = []
+  var ast = doc.ast
+  var _i = 0
+  var _j = 0
+  var stmt = null
+  var decl = null
+  var name = null
+  var kind = null
+  var range = null
+
+  if (ast == null || ast.statements == null) {
+    return symbols
+  }
+
+  while (_i < length(ast.statements)) {
+    stmt = ast.statements[_i]
+
+    if (stmt.kind == "var" || stmt.kind == "def") {
+      name = null
+      kind = KIND_VARIABLE
+
+      if (stmt.left != null && stmt.left.name != null) {
+        name = stmt.left.name
+      }
+
+      if (stmt.kind == "def") {
+        kind = KIND_CONSTANT
+      }
+
+      if (stmt.right != null && (stmt.right.kind == "function" || stmt.right.kind == "arrow function")) {
+        kind = KIND_FUNCTION
+      }
+
+      if (name != null) {
+        range = {
+          start: {line: stmt.from_row, character: stmt.from_column},
+          end: {line: stmt.to_row, character: stmt.to_column}
+        }
+        symbols[] = {
+          name: name,
+          kind: kind,
+          range: range,
+          selectionRange: {
+            start: {line: stmt.left.from_row, character: stmt.left.from_column},
+            end: {line: stmt.left.to_row, character: stmt.left.to_column}
+          }
+        }
+      }
+    }
+
+    if (stmt.kind == "var_list" && stmt.list != null) {
+      _j = 0
+      while (_j < length(stmt.list)) {
+        decl = stmt.list[_j]
+        if (decl.left != null && decl.left.name != null) {
+          kind = (decl.kind == "def") ? KIND_CONSTANT : KIND_VARIABLE
+          if (decl.right != null && (decl.right.kind == "function" || decl.right.kind == "arrow function")) {
+            kind = KIND_FUNCTION
+          }
+          range = {
+            start: {line: decl.from_row, character: decl.from_column},
+            end: {line: decl.to_row, character: decl.to_column}
+          }
+          symbols[] = {
+            name: decl.left.name,
+            kind: kind,
+            range: range,
+            selectionRange: {
+              start: {line: decl.left.from_row, character: decl.left.from_column},
+              end: {line: decl.left.to_row, character: decl.left.to_column}
+            }
+          }
+        }
+        _j = _j + 1
+      }
+    }
+
+    _i = _i + 1
+  }
+
+  return symbols
+}
+
+// Find the declaration location of a name at a given position.
+var definition = function(doc, line, col, token_at) {
+  var tok = token_at(doc, line, col)
+  var ast = doc.ast
+  var name = null
+  var _i = 0
+  var _j = 0
+  var scope = null
+  var v = null
+  var decl = null
+
+  if (tok == null || tok.kind != "name" || tok.value == null) {
+    return null
+  }
+
+  name = tok.value
+
+  if (ast == null) {
+    return null
+  }
+
+  // Search through scopes for the variable declaration
+  if (ast.scopes != null) {
+    _i = 0
+    while (_i < length(ast.scopes)) {
+      scope = ast.scopes[_i]
+      if (scope.vars != null) {
+        _j = 0
+        while (_j < length(scope.vars)) {
+          v = scope.vars[_j]
+          if (v.name == name) {
+            decl = find_declaration(ast.statements, name)
+            if (decl != null) {
+              return {
+                uri: doc.uri,
+                range: {
+                  start: {line: decl.from_row, character: decl.from_column},
+                  end: {line: decl.to_row, character: decl.to_column}
+                }
+              }
+            }
+          }
+          _j = _j + 1
+        }
+      }
+      _i = _i + 1
+    }
+  }
+
+  // Fallback: walk statements for var/def with this name
+  decl = find_declaration(ast.statements, name)
+  if (decl != null) {
+    return {
+      uri: doc.uri,
+      range: {
+        start: {line: decl.from_row, character: decl.from_column},
+        end: {line: decl.to_row, character: decl.to_column}
+      }
+    }
+  }
+
+  return null
+}
+
+// Recursively search statements for a var/def declaration of a given name.
+var find_declaration = function(statements, name) {
+  var _i = 0
+  var _j = 0
+  var stmt = null
+  var result = null
+
+  if (statements == null) {
+    return null
+  }
+
+  while (_i < length(statements)) {
+    stmt = statements[_i]
+
+    // Direct var/def
+    if ((stmt.kind == "var" || stmt.kind == "def")
+        && stmt.left != null && stmt.left.name == name) {
+      return stmt
+    }
+
+    // var_list
+    if (stmt.kind == "var_list" && stmt.list != null) {
+      _j = 0
+      while (_j < length(stmt.list)) {
+        if (stmt.list[_j].left != null && stmt.list[_j].left.name == name) {
+          return stmt.list[_j]
+        }
+        _j = _j + 1
+      }
+    }
+
+    // Recurse into blocks
+    if (stmt.statements != null) {
+      result = find_declaration(stmt.statements, name)
+      if (result != null) {
+        return result
+      }
+    }
+
+    // if/else
+    if (stmt.kind == "if") {
+      if (stmt.then != null && stmt.then.statements != null) {
+        result = find_declaration(stmt.then.statements, name)
+        if (result != null) {
+          return result
+        }
+      }
+      if (stmt.else != null && stmt.else.statements != null) {
+        result = find_declaration(stmt.else.statements, name)
+        if (result != null) {
+          return result
+        }
+      }
+    }
+
+    // Function body
+    if ((stmt.kind == "function" || stmt.kind == "arrow function") && stmt.statements != null) {
+      result = find_declaration(stmt.statements, name)
+      if (result != null) {
+        return result
+      }
+    }
+
+    // var/def with function right side
+    if ((stmt.kind == "var" || stmt.kind == "def") && stmt.right != null) {
+      if ((stmt.right.kind == "function" || stmt.right.kind == "arrow function") && stmt.right.statements != null) {
+        result = find_declaration(stmt.right.statements, name)
+        if (result != null) {
+          return result
+        }
+      }
+    }
+
+    _i = _i + 1
+  }
+  return null
+}
+
+return {
+  document_symbols: document_symbols,
+  definition: definition
+}

editors/vscode/package.json

@@ -0,0 +1,62 @@
+{
+  "name": "pit-language",
+  "displayName": "ƿit Language",
+  "description": "Language support for ƿit (.ce/.cm) — syntax highlighting, diagnostics, completions, hover, and go-to-definition",
+  "version": "0.1.0",
+  "publisher": "pit-lang",
+  "engines": {
+    "vscode": "^1.75.0"
+  },
+  "categories": [
+    "Programming Languages"
+  ],
+  "activationEvents": [
+    "onLanguage:pit"
+  ],
+  "main": "./out/extension.js",
+  "contributes": {
+    "languages": [
+      {
+        "id": "pit",
+        "aliases": [
+          "ƿit",
+          "pit"
+        ],
+        "extensions": [
+          ".ce",
+          ".cm"
+        ],
+        "configuration": "./language-configuration.json"
+      }
+    ],
+    "grammars": [
+      {
+        "language": "pit",
+        "scopeName": "source.pit",
+        "path": "./syntaxes/pit.tmLanguage.json"
+      }
+    ],
+    "configuration": {
+      "title": "ƿit",
+      "properties": {
+        "pit.cellPath": {
+          "type": "string",
+          "default": "cell",
+          "description": "Path to the cell executable"
+        }
+      }
+    }
+  },
+  "scripts": {
+    "compile": "tsc -p ./",
+    "watch": "tsc -watch -p ./"
+  },
+  "dependencies": {
+    "vscode-languageclient": "^9.0.0",
+    "vscode-languageserver-protocol": "^3.17.0"
+  },
+  "devDependencies": {
+    "@types/vscode": "^1.75.0",
+    "typescript": "^5.0.0"
+  }
+}

editors/vscode/src/extension.ts

@@ -0,0 +1,44 @@
+import * as path from "path";
+import { workspace, ExtensionContext } from "vscode";
+import {
+  LanguageClient,
+  LanguageClientOptions,
+  ServerOptions,
+} from "vscode-languageclient/node";
+
+let client: LanguageClient;
+
+export function activate(context: ExtensionContext) {
+  const config = workspace.getConfiguration("pit");
+  const cellPath = config.get<string>("cellPath", "cell");
+  const lspDir = path.join(context.extensionPath, "lsp");
+
+  const serverOptions: ServerOptions = {
+    command: cellPath,
+    args: ["lsp/lsp"],
+    options: { cwd: lspDir },
+  };
+
+  const clientOptions: LanguageClientOptions = {
+    documentSelector: [{ scheme: "file", language: "pit" }],
+    synchronize: {
+      fileEvents: workspace.createFileSystemWatcher("**/*.{ce,cm}"),
+    },
+  };
+
+  client = new LanguageClient(
+    "pitLanguageServer",
+    "ƿit Language Server",
+    serverOptions,
+    clientOptions
+  );
+
+  client.start();
+}
+
+export function deactivate(): Thenable<void> | undefined {
+  if (!client) {
+    return undefined;
+  }
+  return client.stop();
+}

editors/vscode/syntaxes/pit.tmLanguage.json

@@ -0,0 +1,160 @@
+{
+  "$schema": "https://raw.githubusercontent.com/martinring/tmlanguage/master/tmlanguage.json",
+  "name": "pit",
+  "scopeName": "source.pit",
+  "patterns": [
+    { "include": "#comment-line" },
+    { "include": "#comment-block" },
+    { "include": "#string-template" },
+    { "include": "#string-double" },
+    { "include": "#regexp" },
+    { "include": "#keyword-control" },
+    { "include": "#keyword-error" },
+    { "include": "#storage-type" },
+    { "include": "#constant-language" },
+    { "include": "#variable-language" },
+    { "include": "#actor-intrinsic" },
+    { "include": "#keyword-operator" },
+    { "include": "#arrow-function" },
+    { "include": "#support-function" },
+    { "include": "#constant-numeric-hex" },
+    { "include": "#constant-numeric-binary" },
+    { "include": "#constant-numeric-octal" },
+    { "include": "#constant-numeric" },
+    { "include": "#punctuation" }
+  ],
+  "repository": {
+    "comment-line": {
+      "name": "comment.line.double-slash.pit",
+      "match": "//.*$"
+    },
+    "comment-block": {
+      "name": "comment.block.pit",
+      "begin": "/\\*",
+      "end": "\\*/",
+      "beginCaptures": { "0": { "name": "punctuation.definition.comment.begin.pit" } },
+      "endCaptures": { "0": { "name": "punctuation.definition.comment.end.pit" } }
+    },
+    "string-double": {
+      "name": "string.quoted.double.pit",
+      "begin": "\"",
+      "end": "\"",
+      "beginCaptures": { "0": { "name": "punctuation.definition.string.begin.pit" } },
+      "endCaptures": { "0": { "name": "punctuation.definition.string.end.pit" } },
+      "patterns": [
+        {
+          "name": "constant.character.escape.pit",
+          "match": "\\\\(?:[\"\\\\bfnrt/]|u[0-9a-fA-F]{4})"
+        }
+      ]
+    },
+    "string-template": {
+      "name": "string.template.pit",
+      "begin": "`",
+      "end": "`",
+      "beginCaptures": { "0": { "name": "punctuation.definition.string.template.begin.pit" } },
+      "endCaptures": { "0": { "name": "punctuation.definition.string.template.end.pit" } },
+      "patterns": [
+        {
+          "name": "constant.character.escape.pit",
+          "match": "\\\\(?:[`\\\\bfnrt/$]|u[0-9a-fA-F]{4})"
+        },
+        {
+          "name": "meta.template.expression.pit",
+          "begin": "\\$\\{",
+          "end": "\\}",
+          "beginCaptures": { "0": { "name": "punctuation.definition.template-expression.begin.pit" } },
+          "endCaptures": { "0": { "name": "punctuation.definition.template-expression.end.pit" } },
+          "patterns": [
+            { "include": "source.pit" }
+          ]
+        }
+      ]
+    },
+    "regexp": {
+      "name": "string.regexp.pit",
+      "begin": "(?<=[=(:,;!&|?~^>]|^|return|disrupt)\\s*(/(?![/*]))",
+      "end": "/([gimsuvy]*)",
+      "beginCaptures": { "1": { "name": "punctuation.definition.string.begin.pit" } },
+      "endCaptures": { "1": { "name": "keyword.other.pit" } },
+      "patterns": [
+        {
+          "name": "constant.character.escape.pit",
+          "match": "\\\\."
+        }
+      ]
+    },
+    "keyword-control": {
+      "name": "keyword.control.pit",
+      "match": "\\b(if|else|for|while|do|break|continue|return|go)\\b"
+    },
+    "keyword-error": {
+      "name": "keyword.control.error.pit",
+      "match": "\\b(disrupt|disruption)\\b"
+    },
+    "storage-type": {
+      "patterns": [
+        {
+          "name": "storage.type.pit",
+          "match": "\\b(var|def)\\b"
+        },
+        {
+          "name": "storage.type.function.pit",
+          "match": "\\bfunction\\b"
+        }
+      ]
+    },
+    "constant-language": {
+      "name": "constant.language.pit",
+      "match": "\\b(null|true|false)\\b"
+    },
+    "variable-language": {
+      "name": "variable.language.this.pit",
+      "match": "\\bthis\\b"
+    },
+    "actor-intrinsic": {
+      "name": "variable.language.actor.pit",
+      "match": "\\$[a-zA-Z_][a-zA-Z0-9_]*"
+    },
+    "keyword-operator": {
+      "name": "keyword.operator.pit",
+      "match": "\\b(delete|in|typeof)\\b"
+    },
+    "arrow-function": {
+      "name": "storage.type.function.arrow.pit",
+      "match": "=>"
+    },
+    "support-function": {
+      "name": "support.function.pit",
+      "match": "\\b(abs|apply|array|ceiling|character|codepoint|ends_with|every|extract|fallback|filter|find|floor|for|format|fraction|is_array|is_blob|is_character|is_data|is_digit|is_false|is_fit|is_function|is_integer|is_letter|is_logical|is_lower|is_null|is_number|is_object|is_pattern|is_stone|is_text|is_true|is_upper|is_whitespace|length|logical|lower|max|min|modulo|neg|normalize|not|number|parallel|print|race|record|reduce|remainder|replace|reverse|round|search|sequence|sign|some|sort|starts_with|stone|text|trim|trunc|upper|whole|meme|proto|isa|splat|use)(?=\\s*\\()"
+    },
+    "constant-numeric-hex": {
+      "name": "constant.numeric.hex.pit",
+      "match": "\\b0[xX][0-9a-fA-F]+\\b"
+    },
+    "constant-numeric-binary": {
+      "name": "constant.numeric.binary.pit",
+      "match": "\\b0[bB][01]+\\b"
+    },
+    "constant-numeric-octal": {
+      "name": "constant.numeric.octal.pit",
+      "match": "\\b0[oO][0-7]+\\b"
+    },
+    "constant-numeric": {
+      "name": "constant.numeric.pit",
+      "match": "\\b[0-9]+(\\.[0-9]+)?([eE][+-]?[0-9]+)?\\b"
+    },
+    "punctuation": {
+      "patterns": [
+        {
+          "name": "punctuation.separator.comma.pit",
+          "match": ","
+        },
+        {
+          "name": "punctuation.terminator.statement.pit",
+          "match": ";"
+        }
+      ]
+    }
+  }
+}

editors/vscode/tsconfig.json

@@ -0,0 +1,13 @@
+{
+  "compilerOptions": {
+    "module": "commonjs",
+    "target": "ES2020",
+    "outDir": "out",
+    "lib": ["ES2020"],
+    "sourceMap": true,
+    "rootDir": "src",
+    "strict": true
+  },
+  "include": ["src"],
+  "exclude": ["node_modules", "out"]
+}

fold.ce

@@ -0,0 +1,13 @@
+var fd = use("fd")
+var json = use("json")
+
+var filename = args[0]
+var src = text(fd.slurp(filename))
+var tokenize = use("tokenize")
+var parse = use("parse")
+var fold = use("fold")
+
+var tok_result = tokenize(src, filename)
+var ast = parse(tok_result.tokens, src, filename, tokenize)
+var folded = fold(ast)
+print(json.encode(folded))

fold.cm

@@ -0,0 +1,968 @@
+// fold.cm — AST optimization pass
+// Constant folding, constant propagation, dead code elimination
+
+var fold = function(ast) {
+  var scopes = ast.scopes
+  var nr_scopes = length(scopes)
+
+  // ============================================================
+  // Helpers
+  // ============================================================
+
+  var is_literal = function(expr) {
+    if (expr == null) return false
+    var k = expr.kind
+    return k == "number" || k == "text" || k == "true" || k == "false" || k == "null"
+  }
+
+  var is_pure = function(expr) {
+    if (expr == null) return true
+    var k = expr.kind
+    var i = 0
+    if (k == "number" || k == "text" || k == "true" || k == "false" ||
+        k == "null" || k == "name" || k == "this") return true
+    if (k == "function") return true
+    if (k == "!" || k == "~" || k == "-unary" || k == "+unary") {
+      return is_pure(expr.expression)
+    }
+    if (k == "array") {
+      i = 0
+      while (i < length(expr.list)) {
+        if (!is_pure(expr.list[i])) return false
+        i = i + 1
+      }
+      return true
+    }
+    if (k == "record") {
+      i = 0
+      while (i < length(expr.list)) {
+        if (!is_pure(expr.list[i].right)) return false
+        i = i + 1
+      }
+      return true
+    }
+    if (k == "then") {
+      return is_pure(expr.expression) && is_pure(expr.then) && is_pure(expr.else)
+    }
+    if (k == "==" || k == "!=" || k == "&&" || k == "||") {
+      return is_pure(expr.left) && is_pure(expr.right)
+    }
+    return false
+  }
+
+  var copy_loc = function(from, to) {
+    to.at = from.at
+    to.from_row = from.from_row
+    to.from_column = from.from_column
+    to.to_row = from.to_row
+    to.to_column = from.to_column
+    return to
+  }
+
+  var make_number = function(val, src) {
+    return copy_loc(src, {kind: "number", value: text(val), number: val})
+  }
+
+  var make_text = function(val, src) {
+    return copy_loc(src, {kind: "text", value: val})
+  }
+
+  var make_bool = function(val, src) {
+    if (val) return copy_loc(src, {kind: "true"})
+    return copy_loc(src, {kind: "false"})
+  }
+
+  var make_null = function(src) {
+    return copy_loc(src, {kind: "null"})
+  }
+
+  var is_truthy_literal = function(expr) {
+    if (expr == null) return null
+    var k = expr.kind
+    var nv = null
+    if (k == "true") return true
+    if (k == "false" || k == "null") return false
+    if (k == "number") {
+      nv = expr.number
+      if (nv == null) nv = number(expr.value)
+      return nv != 0
+    }
+    if (k == "text") return length(expr.value) > 0
+    return null
+  }
+
+  // ============================================================
+  // Scope helpers
+  // ============================================================
+
+  var find_scope = function(fn_nr) {
+    var i = 0
+    while (i < nr_scopes) {
+      if (scopes[i].function_nr == fn_nr) return scopes[i]
+      i = i + 1
+    }
+    return null
+  }
+
+  var scope_var = function(fn_nr, name) {
+    var sc = find_scope(fn_nr)
+    if (sc == null) return null
+    return sc[name]
+  }
+
+  var remove_scope_var = function(fn_nr, name) {
+    var sc = find_scope(fn_nr)
+    if (sc == null) return null
+    delete sc[name]
+  }
+
+  // ============================================================
+  // Pass 1: pre-scan for constants and function arities
+  // ============================================================
+
+  var const_defs = {}
+  var fn_arities = {}
+
+  var register_const = function(fn_nr, name, lit_node) {
+    var key = text(fn_nr)
+    if (const_defs[key] == null) const_defs[key] = {}
+    const_defs[key][name] = lit_node
+  }
+
+  var get_const = function(fn_nr, name) {
+    var key = text(fn_nr)
+    if (const_defs[key] == null) return null
+    return const_defs[key][name]
+  }
+
+  var register_arity = function(fn_nr, name, count) {
+    var key = text(fn_nr)
+    if (fn_arities[key] == null) fn_arities[key] = {}
+    fn_arities[key][name] = count
+  }
+
+  var pre_scan_stmts = null
+  var pre_scan_fn = null
+
+  pre_scan_fn = function(node) {
+    if (node == null) return null
+    if (node.statements != null) pre_scan_stmts(node.statements, node.function_nr)
+    if (node.disruption != null) pre_scan_stmts(node.disruption, node.function_nr)
+  }
+
+  pre_scan_stmts = function(stmts, fn_nr) {
+    var i = 0
+    var j = 0
+    var stmt = null
+    var kind = null
+    var name = null
+    var sv = null
+    var item = null
+    while (i < length(stmts)) {
+      stmt = stmts[i]
+      kind = stmt.kind
+      if (kind == "def") {
+        name = stmt.left.name
+        if (name != null && is_literal(stmt.right)) {
+          sv = scope_var(fn_nr, name)
+          if (sv != null && !sv.closure) {
+            register_const(fn_nr, name, stmt.right)
+          }
+        }
+      } else if (kind == "function") {
+        name = stmt.name
+        if (name != null && stmt.arity != null) {
+          register_arity(fn_nr, name, stmt.arity)
+        }
+        pre_scan_fn(stmt)
+      } else if (kind == "var") {
+        if (stmt.right != null && stmt.right.kind == "function" && stmt.right.arity != null) {
+          name = stmt.left.name
+          if (name != null) {
+            sv = scope_var(fn_nr, name)
+            if (sv != null && sv.make == "var") {
+              register_arity(fn_nr, name, stmt.right.arity)
+            }
+          }
+        }
+      } else if (kind == "var_list") {
+        j = 0
+        while (j < length(stmt.list)) {
+          item = stmt.list[j]
+          if (item.kind == "var" && item.right != null && item.right.kind == "function" && item.right.arity != null) {
+            name = item.left.name
+            if (name != null) {
+              sv = scope_var(fn_nr, name)
+              if (sv != null && sv.make == "var") {
+                register_arity(fn_nr, name, item.right.arity)
+              }
+            }
+          }
+          j = j + 1
+        }
+      }
+      i = i + 1
+    }
+  }
+
+  var pre_scan_expr_fns = null
+  pre_scan_expr_fns = function(expr) {
+    if (expr == null) return null
+    var k = expr.kind
+    var i = 0
+    if (k == "function") {
+      pre_scan_fn(expr)
+    }
+    if (expr.left != null) pre_scan_expr_fns(expr.left)
+    if (expr.right != null) pre_scan_expr_fns(expr.right)
+    if (expr.expression != null) pre_scan_expr_fns(expr.expression)
+    if (expr.then != null) pre_scan_expr_fns(expr.then)
+    if (expr.else != null) pre_scan_expr_fns(expr.else)
+    if (k == "(" || k == "array") {
+      i = 0
+      while (i < length(expr.list)) {
+        pre_scan_expr_fns(expr.list[i])
+        i = i + 1
+      }
+    }
+    if (k == "record") {
+      i = 0
+      while (i < length(expr.list)) {
+        pre_scan_expr_fns(expr.list[i].right)
+        i = i + 1
+      }
+    }
+  }
+
+  var pre_scan_stmt_exprs = null
+  pre_scan_stmt_exprs = function(stmts, fn_nr) {
+    var i = 0
+    var j = 0
+    var stmt = null
+    var kind = null
+    while (i < length(stmts)) {
+      stmt = stmts[i]
+      kind = stmt.kind
+      if (kind == "var" || kind == "def") {
+        pre_scan_expr_fns(stmt.right)
+      } else if (kind == "var_list") {
+        j = 0
+        while (j < length(stmt.list)) {
+          pre_scan_expr_fns(stmt.list[j].right)
+          j = j + 1
+        }
+      } else if (kind == "call") {
+        pre_scan_expr_fns(stmt.expression)
+      } else if (kind == "if") {
+        pre_scan_expr_fns(stmt.expression)
+        pre_scan_stmt_exprs(stmt.then, fn_nr)
+        pre_scan_stmt_exprs(stmt.list, fn_nr)
+        if (stmt.else != null) pre_scan_stmt_exprs(stmt.else, fn_nr)
+      } else if (kind == "while" || kind == "do") {
+        pre_scan_expr_fns(stmt.expression)
+        pre_scan_stmt_exprs(stmt.statements, fn_nr)
+      } else if (kind == "for") {
+        if (stmt.init != null) {
+          if (stmt.init.kind == "var" || stmt.init.kind == "def") {
+            pre_scan_expr_fns(stmt.init.right)
+          } else {
+            pre_scan_expr_fns(stmt.init)
+          }
+        }
+        pre_scan_expr_fns(stmt.test)
+        pre_scan_expr_fns(stmt.update)
+        pre_scan_stmt_exprs(stmt.statements, fn_nr)
+      } else if (kind == "return" || kind == "go") {
+        pre_scan_expr_fns(stmt.expression)
+      } else if (kind == "block") {
+        pre_scan_stmt_exprs(stmt.statements, fn_nr)
+      } else if (kind == "label") {
+        if (stmt.statement != null) {
+          pre_scan_stmt_exprs([stmt.statement], fn_nr)
+        }
+      } else if (kind == "function") {
+        // already handled in pre_scan_stmts
+        null
+      }
+      i = i + 1
+    }
+  }
+
+  var pre_scan = function() {
+    pre_scan_stmts(ast.statements, 0)
+    pre_scan_stmts(ast.functions, 0)
+    pre_scan_stmt_exprs(ast.statements, 0)
+    pre_scan_stmt_exprs(ast.functions, 0)
+  }
+
+  // ============================================================
+  // Pass 2: fold expressions and statements
+  // ============================================================
+
+  var fold_expr = null
+  var fold_stmt = null
+  var fold_stmts = null
+
+  fold_expr = function(expr, fn_nr) {
+    if (expr == null) return null
+    var k = expr.kind
+    var left = null
+    var right = null
+    var lv = null
+    var rv = null
+    var result = null
+    var i = 0
+    var sv = null
+    var lit = null
+    var cond_k = null
+    var ek = null
+    var target = null
+    var ar = null
+    var akey = null
+    var tv = null
+
+    // Recurse into children first (bottom-up)
+    if (k == "+" || k == "-" || k == "*" || k == "/" || k == "%" ||
+        k == "**" || k == "==" || k == "!=" || k == "<" || k == ">" ||
+        k == "<=" || k == ">=" || k == "&" || k == "|" || k == "^" ||
+        k == "<<" || k == ">>" || k == ">>>" || k == "&&" || k == "||" ||
+        k == "," || k == "in") {
+      expr.left = fold_expr(expr.left, fn_nr)
+      expr.right = fold_expr(expr.right, fn_nr)
+    } else if (k == "." || k == "[") {
+      expr.left = fold_expr(expr.left, fn_nr)
+      if (k == "[" && expr.right != null) expr.right = fold_expr(expr.right, fn_nr)
+    } else if (k == "!" || k == "~" || k == "-unary" || k == "+unary" || k == "delete") {
+      expr.expression = fold_expr(expr.expression, fn_nr)
+    } else if (k == "++" || k == "--") {
+      return expr
+    } else if (k == "then") {
+      expr.expression = fold_expr(expr.expression, fn_nr)
+      expr.then = fold_expr(expr.then, fn_nr)
+      expr.else = fold_expr(expr.else, fn_nr)
+    } else if (k == "(") {
+      expr.expression = fold_expr(expr.expression, fn_nr)
+      i = 0
+      while (i < length(expr.list)) {
+        expr.list[i] = fold_expr(expr.list[i], fn_nr)
+        i = i + 1
+      }
+    } else if (k == "array") {
+      i = 0
+      while (i < length(expr.list)) {
+        expr.list[i] = fold_expr(expr.list[i], fn_nr)
+        i = i + 1
+      }
+    } else if (k == "record") {
+      i = 0
+      while (i < length(expr.list)) {
+        expr.list[i].right = fold_expr(expr.list[i].right, fn_nr)
+        i = i + 1
+      }
+    } else if (k == "text literal") {
+      i = 0
+      while (i < length(expr.list)) {
+        expr.list[i] = fold_expr(expr.list[i], fn_nr)
+        i = i + 1
+      }
+    } else if (k == "function") {
+      fold_fn(expr)
+      return expr
+    } else if (k == "assign" || k == "+=" || k == "-=" || k == "*=" ||
+               k == "/=" || k == "%=" || k == "<<=" || k == ">>=" ||
+               k == ">>>=" || k == "&=" || k == "^=" || k == "|=" ||
+               k == "**=" || k == "&&=" || k == "||=") {
+      expr.right = fold_expr(expr.right, fn_nr)
+      return expr
+    }
+
+    // Constant propagation: name → literal
+    if (k == "name" && expr.level == 0) {
+      lit = get_const(fn_nr, expr.name)
+      if (lit != null) {
+        sv = scope_var(fn_nr, expr.name)
+        if (sv != null && !sv.closure) {
+          return copy_loc(expr, {kind: lit.kind, value: lit.value, number: lit.number})
+        }
+      }
+      return expr
+    }
+
+    // Binary constant folding
+    if (k == "+" || k == "-" || k == "*" || k == "/" || k == "%" || k == "**") {
+      left = expr.left
+      right = expr.right
+      if (left != null && right != null && left.kind == "number" && right.kind == "number") {
+        lv = left.number
+        rv = right.number
+        if (lv == null) lv = number(left.value)
+        if (rv == null) rv = number(right.value)
+        if (k == "/") {
+          if (rv == 0) return make_null(expr)
+        }
+        if (k == "%") {
+          if (rv == 0) return make_null(expr)
+        }
+        result = null
+        if (k == "+") result = lv + rv
+        else if (k == "-") result = lv - rv
+        else if (k == "*") result = lv * rv
+        else if (k == "/") result = lv / rv
+        else if (k == "%") result = lv % rv
+        else if (k == "**") result = lv ** rv
+        if (result == null) return make_null(expr)
+        return make_number(result, expr)
+      }
+      // text + text
+      if (k == "+" && left != null && right != null && left.kind == "text" && right.kind == "text") {
+        return make_text(left.value + right.value, expr)
+      }
+      return expr
+    }
+
+    // Comparison folding
+    if (k == "==" || k == "!=" || k == "<" || k == ">" || k == "<=" || k == ">=") {
+      left = expr.left
+      right = expr.right
+      if (left != null && right != null) {
+        if (left.kind == "number" && right.kind == "number") {
+          lv = left.number
+          rv = right.number
+          if (lv == null) lv = number(left.value)
+          if (rv == null) rv = number(right.value)
+          if (k == "==") return make_bool(lv == rv, expr)
+          if (k == "!=") return make_bool(lv != rv, expr)
+          if (k == "<") return make_bool(lv < rv, expr)
+          if (k == ">") return make_bool(lv > rv, expr)
+          if (k == "<=") return make_bool(lv <= rv, expr)
+          if (k == ">=") return make_bool(lv >= rv, expr)
+        }
+        if (left.kind == "text" && right.kind == "text") {
+          if (k == "==") return make_bool(left.value == right.value, expr)
+          if (k == "!=") return make_bool(left.value != right.value, expr)
+        }
+      }
+      return expr
+    }
+
+    // Bitwise folding
+    if (k == "&" || k == "|" || k == "^" || k == "<<" || k == ">>") {
+      left = expr.left
+      right = expr.right
+      if (left != null && right != null && left.kind == "number" && right.kind == "number") {
+        lv = left.number
+        rv = right.number
+        if (lv == null) lv = number(left.value)
+        if (rv == null) rv = number(right.value)
+        if (k == "&") return make_number(lv & rv, expr)
+        if (k == "|") return make_number(lv | rv, expr)
+        if (k == "^") return make_number(lv ^ rv, expr)
+        if (k == "<<") return make_number(lv << rv, expr)
+        if (k == ">>") return make_number(lv >> rv, expr)
+      }
+      return expr
+    }
+
+    // Unary folding
+    if (k == "!") {
+      if (expr.expression != null) {
+        ek = expr.expression.kind
+        if (ek == "true") return make_bool(false, expr)
+        if (ek == "false") return make_bool(true, expr)
+      }
+      return expr
+    }
+    if (k == "~") {
+      if (expr.expression != null && expr.expression.kind == "number") {
+        lv = expr.expression.number
+        if (lv == null) lv = number(expr.expression.value)
+        return make_number(~lv, expr)
+      }
+      return expr
+    }
+    if (k == "-unary") {
+      if (expr.expression != null && expr.expression.kind == "number") {
+        lv = expr.expression.number
+        if (lv == null) lv = number(expr.expression.value)
+        return make_number(0 - lv, expr)
+      }
+      return expr
+    }
+
+    // Ternary with literal condition
+    if (k == "then") {
+      tv = is_truthy_literal(expr.expression)
+      if (tv == true) return expr.then
+      if (tv == false) return expr.else
+      return expr
+    }
+
+    // Call: stamp arity
+    if (k == "(") {
+      target = expr.expression
+      if (target != null && target.kind == "name" && target.level == 0) {
+        ar = null
+        akey = text(fn_nr)
+        if (fn_arities[akey] != null) ar = fn_arities[akey][target.name]
+        if (ar != null) expr.arity = ar
+      }
+      return expr
+    }
+
+    return expr
+  }
+
+  var fold_fn = null
+
+  fold_stmt = function(stmt, fn_nr) {
+    if (stmt == null) return null
+    var k = stmt.kind
+    var i = 0
+    var sv = null
+    var cond_k = null
+    var ik = null
+    var tv = null
+
+    if (k == "var" || k == "def") {
+      stmt.right = fold_expr(stmt.right, fn_nr)
+      return stmt
+    }
+    if (k == "var_list") {
+      i = 0
+      while (i < length(stmt.list)) {
+        stmt.list[i] = fold_stmt(stmt.list[i], fn_nr)
+        i = i + 1
+      }
+      return stmt
+    }
+    if (k == "call") {
+      stmt.expression = fold_expr(stmt.expression, fn_nr)
+      return stmt
+    }
+    if (k == "if") {
+      stmt.expression = fold_expr(stmt.expression, fn_nr)
+      tv = is_truthy_literal(stmt.expression)
+      if (tv == true) {
+        stmt.then = fold_stmts(stmt.then, fn_nr)
+        return {kind: "block", statements: stmt.then,
+          at: stmt.at, from_row: stmt.from_row, from_column: stmt.from_column,
+          to_row: stmt.to_row, to_column: stmt.to_column}
+      }
+      if (tv == false) {
+        if (stmt.else != null && length(stmt.else) > 0) {
+          stmt.else = fold_stmts(stmt.else, fn_nr)
+          return {kind: "block", statements: stmt.else,
+            at: stmt.at, from_row: stmt.from_row, from_column: stmt.from_column,
+            to_row: stmt.to_row, to_column: stmt.to_column}
+        }
+        if (stmt.list != null && length(stmt.list) > 0) {
+          return fold_stmt(stmt.list[0], fn_nr)
+        }
+        return null
+      }
+      stmt.then = fold_stmts(stmt.then, fn_nr)
+      stmt.list = fold_stmts(stmt.list, fn_nr)
+      if (stmt.else != null) stmt.else = fold_stmts(stmt.else, fn_nr)
+      return stmt
+    }
+    if (k == "while") {
+      stmt.expression = fold_expr(stmt.expression, fn_nr)
+      if (stmt.expression.kind == "false" || stmt.expression.kind == "null") return null
+      stmt.statements = fold_stmts(stmt.statements, fn_nr)
+      return stmt
+    }
+    if (k == "do") {
+      stmt.statements = fold_stmts(stmt.statements, fn_nr)
+      stmt.expression = fold_expr(stmt.expression, fn_nr)
+      return stmt
+    }
+    if (k == "for") {
+      if (stmt.init != null) {
+        ik = stmt.init.kind
+        if (ik == "var" || ik == "def") {
+          stmt.init = fold_stmt(stmt.init, fn_nr)
+        } else {
+          stmt.init = fold_expr(stmt.init, fn_nr)
+        }
+      }
+      if (stmt.test != null) stmt.test = fold_expr(stmt.test, fn_nr)
+      if (stmt.update != null) stmt.update = fold_expr(stmt.update, fn_nr)
+      stmt.statements = fold_stmts(stmt.statements, fn_nr)
+      return stmt
+    }
+    if (k == "return" || k == "go") {
+      stmt.expression = fold_expr(stmt.expression, fn_nr)
+      return stmt
+    }
+    if (k == "block") {
+      stmt.statements = fold_stmts(stmt.statements, fn_nr)
+      return stmt
+    }
+    if (k == "label") {
+      stmt.statement = fold_stmt(stmt.statement, fn_nr)
+      return stmt
+    }
+    if (k == "function") {
+      fold_fn(stmt)
+      return stmt
+    }
+    return stmt
+  }
+
+  fold_stmts = function(stmts, fn_nr) {
+    var i = 0
+    var stmt = null
+    var out = []
+    var sv = null
+    var name = null
+    while (i < length(stmts)) {
+      stmt = fold_stmt(stmts[i], fn_nr)
+      if (stmt == null) {
+        i = i + 1
+        continue
+      }
+      // Dead code elimination: unused pure var/def
+      if (stmt.kind == "var" || stmt.kind == "def") {
+        name = stmt.left.name
+        if (name != null) {
+          sv = scope_var(fn_nr, name)
+          if (sv != null && sv.nr_uses == 0 && is_pure(stmt.right)) {
+            stmt.dead = true
+          }
+        }
+      }
+      // Dead function elimination
+      if (stmt.kind == "function" && stmt.name != null) {
+        sv = scope_var(fn_nr, stmt.name)
+        if (sv != null && sv.nr_uses == 0) {
+          stmt.dead = true
+        }
+      }
+      if (stmt.dead != true) push(out, stmt)
+      i = i + 1
+    }
+    return out
+  }
+
+  fold_fn = function(node) {
+    if (node == null) return null
+    var fn_nr = node.function_nr
+    if (fn_nr == null) return null
+    // Fold param defaults
+    var i = 0
+    while (i < length(node.list)) {
+      if (node.list[i].expression != null) {
+        node.list[i].expression = fold_expr(node.list[i].expression, fn_nr)
+      }
+      i = i + 1
+    }
+    if (node.statements != null) node.statements = fold_stmts(node.statements, fn_nr)
+    if (node.disruption != null) node.disruption = fold_stmts(node.disruption, fn_nr)
+  }
+
+  // ============================================================
+  // Pass 3: cleanup scopes
+  // ============================================================
+
+  var cleanup = function() {
+    var i = 0
+    var sc = null
+    var keys = null
+    var j = 0
+    var key = null
+    var entry = null
+    var slots = 0
+    var close_slots = 0
+
+    // Remove dead vars from scope records and recalculate slot counts
+    while (i < nr_scopes) {
+      sc = scopes[i]
+      keys = array(sc)
+      slots = 0
+      close_slots = 0
+      j = 0
+      while (j < length(keys)) {
+        key = keys[j]
+        if (key != "function_nr") {
+          entry = sc[key]
+          if (entry != null && entry.nr_uses == 0 && entry.make != "input" && entry.make != "function") {
+            delete sc[key]
+          } else if (entry != null) {
+            slots = slots + 1
+            if (entry.closure) close_slots = close_slots + 1
+          }
+        }
+        j = j + 1
+      }
+      i = i + 1
+    }
+
+    // Update nr_slots and nr_close_slots on function nodes
+    var update_fn_slots = null
+    update_fn_slots = function(node) {
+      if (node == null) return null
+      var fn_nr = node.function_nr
+      if (fn_nr == null) return null
+      var sc = find_scope(fn_nr)
+      if (sc == null) return null
+      var keys = array(sc)
+      var s = 0
+      var cs = 0
+      var ki = 0
+      var ent = null
+      while (ki < length(keys)) {
+        if (keys[ki] != "function_nr") {
+          ent = sc[keys[ki]]
+          if (ent != null) {
+            s = s + 1
+            if (ent.closure) cs = cs + 1
+          }
+        }
+        ki = ki + 1
+      }
+      node.nr_slots = s
+      node.nr_close_slots = cs
+    }
+
+    var walk_stmts_for_fns = null
+    var walk_expr_for_fns = null
+
+    walk_expr_for_fns = function(expr) {
+      if (expr == null) return null
+      var k = expr.kind
+      var i = 0
+      if (k == "function") {
+        update_fn_slots(expr)
+        walk_stmts_for_fns(expr.statements)
+        walk_stmts_for_fns(expr.disruption)
+        return null
+      }
+      if (expr.left != null) walk_expr_for_fns(expr.left)
+      if (expr.right != null) walk_expr_for_fns(expr.right)
+      if (expr.expression != null) walk_expr_for_fns(expr.expression)
+      if (expr.then != null) walk_expr_for_fns(expr.then)
+      if (expr.else != null) walk_expr_for_fns(expr.else)
+      if (k == "(" || k == "array" || k == "text literal") {
+        i = 0
+        while (i < length(expr.list)) {
+          walk_expr_for_fns(expr.list[i])
+          i = i + 1
+        }
+      }
+      if (k == "record") {
+        i = 0
+        while (i < length(expr.list)) {
+          walk_expr_for_fns(expr.list[i].right)
+          i = i + 1
+        }
+      }
+    }
+
+    walk_stmts_for_fns = function(stmts) {
+      if (stmts == null) return null
+      var i = 0
+      var j = 0
+      var stmt = null
+      var k = null
+      while (i < length(stmts)) {
+        stmt = stmts[i]
+        k = stmt.kind
+        if (k == "function") {
+          update_fn_slots(stmt)
+          walk_stmts_for_fns(stmt.statements)
+          walk_stmts_for_fns(stmt.disruption)
+        } else if (k == "var" || k == "def") {
+          walk_expr_for_fns(stmt.right)
+        } else if (k == "var_list") {
+          j = 0
+          while (j < length(stmt.list)) {
+            walk_expr_for_fns(stmt.list[j].right)
+            j = j + 1
+          }
+        } else if (k == "call") {
+          walk_expr_for_fns(stmt.expression)
+        } else if (k == "if") {
+          walk_expr_for_fns(stmt.expression)
+          walk_stmts_for_fns(stmt.then)
+          walk_stmts_for_fns(stmt.list)
+          if (stmt.else != null) walk_stmts_for_fns(stmt.else)
+        } else if (k == "while" || k == "do") {
+          walk_expr_for_fns(stmt.expression)
+          walk_stmts_for_fns(stmt.statements)
+        } else if (k == "for") {
+          if (stmt.init != null) {
+            if (stmt.init.kind == "var" || stmt.init.kind == "def") {
+              walk_expr_for_fns(stmt.init.right)
+            } else {
+              walk_expr_for_fns(stmt.init)
+            }
+          }
+          walk_expr_for_fns(stmt.test)
+          walk_expr_for_fns(stmt.update)
+          walk_stmts_for_fns(stmt.statements)
+        } else if (k == "return" || k == "go") {
+          walk_expr_for_fns(stmt.expression)
+        } else if (k == "block") {
+          walk_stmts_for_fns(stmt.statements)
+        } else if (k == "label") {
+          if (stmt.statement != null) walk_stmts_for_fns([stmt.statement])
+        }
+        i = i + 1
+      }
+    }
+
+    walk_stmts_for_fns(ast.statements)
+    walk_stmts_for_fns(ast.functions)
+
+    // Update intrinsics: collect what's still referenced
+    var used_intrinsics = {}
+    var collect_intrinsics = null
+    var collect_expr_intrinsics = null
+
+    collect_expr_intrinsics = function(expr) {
+      if (expr == null) return null
+      var k = expr.kind
+      var i = 0
+      if (k == "name" && expr.level == -1 && expr.name != null && expr.make != "functino") {
+        used_intrinsics[expr.name] = true
+      }
+      if (expr.left != null) collect_expr_intrinsics(expr.left)
+      if (expr.right != null) collect_expr_intrinsics(expr.right)
+      if (expr.expression != null) collect_expr_intrinsics(expr.expression)
+      if (expr.then != null) collect_expr_intrinsics(expr.then)
+      if (expr.else != null) collect_expr_intrinsics(expr.else)
+      if (k == "(" || k == "array" || k == "text literal") {
+        i = 0
+        while (i < length(expr.list)) {
+          collect_expr_intrinsics(expr.list[i])
+          i = i + 1
+        }
+      }
+      if (k == "record") {
+        i = 0
+        while (i < length(expr.list)) {
+          collect_expr_intrinsics(expr.list[i].right)
+          i = i + 1
+        }
+      }
+      if (k == "function") {
+        collect_intrinsics(expr.statements)
+        collect_intrinsics(expr.disruption)
+        i = 0
+        while (i < length(expr.list)) {
+          if (expr.list[i].expression != null) {
+            collect_expr_intrinsics(expr.list[i].expression)
+          }
+          i = i + 1
+        }
+      }
+    }
+
+    collect_intrinsics = function(stmts) {
+      if (stmts == null) return null
+      var i = 0
+      var j = 0
+      var pi = 0
+      var stmt = null
+      var k = null
+      while (i < length(stmts)) {
+        stmt = stmts[i]
+        k = stmt.kind
+        if (k == "var" || k == "def") {
+          collect_expr_intrinsics(stmt.right)
+        } else if (k == "var_list") {
+          j = 0
+          while (j < length(stmt.list)) {
+            collect_expr_intrinsics(stmt.list[j].right)
+            j = j + 1
+          }
+        } else if (k == "call") {
+          collect_expr_intrinsics(stmt.expression)
+        } else if (k == "if") {
+          collect_expr_intrinsics(stmt.expression)
+          collect_intrinsics(stmt.then)
+          collect_intrinsics(stmt.list)
+          if (stmt.else != null) collect_intrinsics(stmt.else)
+        } else if (k == "while" || k == "do") {
+          collect_expr_intrinsics(stmt.expression)
+          collect_intrinsics(stmt.statements)
+        } else if (k == "for") {
+          if (stmt.init != null) {
+            if (stmt.init.kind == "var" || stmt.init.kind == "def") {
+              collect_expr_intrinsics(stmt.init.right)
+            } else {
+              collect_expr_intrinsics(stmt.init)
+            }
+          }
+          collect_expr_intrinsics(stmt.test)
+          collect_expr_intrinsics(stmt.update)
+          collect_intrinsics(stmt.statements)
+        } else if (k == "return" || k == "go") {
+          collect_expr_intrinsics(stmt.expression)
+        } else if (k == "function") {
+          collect_intrinsics(stmt.statements)
+          collect_intrinsics(stmt.disruption)
+          pi = 0
+          while (pi < length(stmt.list)) {
+            if (stmt.list[pi].expression != null) {
+              collect_expr_intrinsics(stmt.list[pi].expression)
+            }
+            pi = pi + 1
+          }
+        } else if (k == "block") {
+          collect_intrinsics(stmt.statements)
+        } else if (k == "label") {
+          if (stmt.statement != null) collect_intrinsics([stmt.statement])
+        }
+        i = i + 1
+      }
+    }
+
+    collect_intrinsics(ast.statements)
+    collect_intrinsics(ast.functions)
+
+    var new_intrinsics = []
+    i = 0
+    while (i < length(ast.intrinsics)) {
+      if (used_intrinsics[ast.intrinsics[i]] == true) {
+        push(new_intrinsics, ast.intrinsics[i])
+      }
+      i = i + 1
+    }
+    ast.intrinsics = new_intrinsics
+  }
+
+  // ============================================================
+  // Main
+  // ============================================================
+
+  pre_scan()
+
+  // Pass 2: fold all statements and functions
+  ast.statements = fold_stmts(ast.statements, 0)
+  var fi = 0
+  while (fi < length(ast.functions)) {
+    fold_fn(ast.functions[fi])
+    fi = fi + 1
+  }
+
+  // Remove dead top-level functions
+  var live_fns = []
+  var fn = null
+  fi = 0
+  while (fi < length(ast.functions)) {
+    fn = ast.functions[fi]
+    if (fn.dead != true) {
+      push(live_fns, fn)
+    }
+    fi = fi + 1
+  }
+  ast.functions = live_fns
+
+  // Pass 3: cleanup
+  cleanup()
+
+  return ast
+}
+
+return fold

internal/bootstrap.cm

@@ -0,0 +1,137 @@
+// Hidden vars (os, args, core_path, use_mcode) come from env
+// args[0] = script name, args[1..] = user args
+var load_internal = os.load_internal
+function use_embed(name) {
+  return load_internal("js_" + name + "_use")
+}
+
+var fd = use_embed('fd')
+var json = use_embed('json')
+
+var use_cache = {}
+use_cache['fd'] = fd
+use_cache['os'] = os
+use_cache['json'] = json
+
+// Bootstrap: load tokenize.cm, parse.cm, fold.cm from pre-compiled mach bytecode
+function use_basic(path) {
+  if (use_cache[path])
+    return use_cache[path]
+  var result = use_embed(replace(path, '/', '_'))
+  use_cache[path] = result
+  return result
+}
+
+// Load a module from .mach bytecode, falling back to .ast.json
+function boot_load(name, env) {
+  var mach_path = name + ".mach"
+  var data = null
+  if (fd.is_file(mach_path)) {
+    data = fd.slurp(mach_path)
+    return mach_load(data, env)
+  }
+  data = text(fd.slurp(name + ".ast.json"))
+  return mach_eval_ast(name, data, env)
+}
+
+var boot_env = {use: use_basic}
+var tokenize_mod = boot_load("tokenize", boot_env)
+var parse_mod = boot_load("parse", boot_env)
+var fold_mod = boot_load("fold", boot_env)
+
+// Optionally load mcode compiler module
+var mcode_mod = null
+if (use_mcode) {
+  mcode_mod = boot_load("mcode", boot_env)
+}
+
+// analyze: tokenize + parse, check for errors
+function analyze(src, filename) {
+  var tok_result = tokenize_mod(src, filename)
+  var ast = parse_mod(tok_result.tokens, src, filename, tokenize_mod)
+  var _i = 0
+  var prev_line = -1
+  var prev_msg = null
+  var e = null
+  var msg = null
+  var line = null
+  var col = null
+  var has_errors = ast.errors != null && length(ast.errors) > 0
+  if (has_errors) {
+    while (_i < length(ast.errors)) {
+      e = ast.errors[_i]
+      msg = e.message
+      line = e.line
+      col = e.column
+      if (msg != prev_msg || line != prev_line) {
+        if (line != null && col != null) {
+          print(`${filename}:${text(line)}:${text(col)}: error: ${msg}`)
+        } else {
+          print(`${filename}: error: ${msg}`)
+        }
+      }
+      prev_line = line
+      prev_msg = msg
+      _i = _i + 1
+    }
+    disrupt
+  }
+  ast = fold_mod(ast)
+  return ast
+}
+
+// Run AST through either mcode or mach pipeline
+function run_ast(name, ast, env) {
+  var compiled = null
+  if (use_mcode) {
+    compiled = mcode_mod(ast)
+    return mcode_run(name, json.encode(compiled), env)
+  }
+  return mach_eval_ast(name, json.encode(ast), env)
+}
+
+// use() with ƿit pipeline for .cm modules
+function use(path) {
+  var file_path = path + '.cm'
+  var script = null
+  var ast = null
+  var result = null
+  if (use_cache[path])
+    return use_cache[path]
+
+  // Check CWD first, then core_path
+  if (!fd.is_file(file_path))
+    file_path = core_path + '/' + path + '.cm'
+
+  if (fd.is_file(file_path)) {
+    script = text(fd.slurp(file_path))
+    ast = analyze(script, file_path)
+    result = run_ast(path, ast, {use: use})
+    use_cache[path] = result
+    return result
+  }
+
+  // Fallback to embedded C module
+  result = use_embed(replace(path, '/', '_'))
+  use_cache[path] = result
+  return result
+}
+
+// Load and run the user's program
+var program = args[0]
+var script_file = program
+
+// Add .ce extension if not already present
+if (!ends_with(script_file, '.ce') && !ends_with(script_file, '.cm'))
+  script_file = program + '.ce'
+
+var user_args = []
+var _j = 1
+while (_j < length(args)) {
+  push(user_args, args[_j])
+  _j = _j + 1
+}
+
+var script = text(fd.slurp(script_file))
+var ast = analyze(script, script_file)
+run_ast(program, ast, {use: use, args: user_args, json: json})

internal/engine.cm

@@ -1,17 +1,21 @@
-(function engine() {
 // Hidden vars (os, actorsym, init, core_path) come from env
 var ACTORDATA = actorsym
 var SYSYM = '__SYSTEM__'
 
 var _cell = {}
+var need_stop = false
 
 var dylib_ext
 
-switch(os.platform()) {
-  case 'Windows': dylib_ext = '.dll'; break;
-  case 'macOS': dylib_ext = '.dylib'; break;
-  case 'Linux': dylib_ext = '.so'; break;
-} 
+var cases = {
+  Windows: '.dll',
+  macOS: '.dylib',
+  Linux: '.so'
+}
+
+print(os.platform())
+
+dylib_ext = cases[os.platform()]
 
 var MOD_EXT = '.cm'
 var ACTOR_EXT = '.ce'  
@@ -51,14 +55,16 @@ var fd = use_embed('fd')
 // Get the shop path from HOME environment
 var home = os.getenv('HOME') || os.getenv('USERPROFILE')
 if (!home) {
-  throw Error('Could not determine home directory')
+  os.print('Could not determine home directory\n')
+  os.exit(1)
 }
 var shop_path = home + '/.cell'
 var packages_path = shop_path + '/packages'
 var core_path = packages_path + '/core'
 
 if (!fd.is_dir(core_path)) {
-  throw Error('Cell shop not found at ' + shop_path + '. Run "cell install" to set up.')
+  os.print('Cell shop not found at ' + shop_path + '. Run "cell install" to set up.\n')
+  os.exit(1)
 }
 
 var use_cache = {}
@@ -79,7 +85,7 @@ function use_core(path) {
     var script_blob = fd.slurp(file_path)
     var script = text(script_blob)
     var mod = `(function setup_module(use){${script}})`
-    var fn = js.eval('core:' + path, mod)
+    var fn = mach_eval('core:' + path, mod)
     var result = call(fn,sym, [use_core])
     use_cache[cache_key] = result;
     return result;
@@ -133,32 +139,27 @@ function log(name, args) {
   var caller = caller_data(1)
   var msg = args[0]
 
-  switch(name) {
-    case 'console':
-      os.print(console_rec(caller.line, caller.file, msg))
-      break
-    case 'error':
-      msg = msg ?? Error()
-      if (is_proto(msg, Error))
-        msg = msg.name + ": " + msg.message + "\n" + msg.stack
-      os.print(console_rec(caller.line, caller.file, msg))
-      break
-    case 'system':
-      msg = "[SYSTEM] " + msg
-      os.print(console_rec(caller.line, caller.file, msg))
-      break
-    default:
-      log.console(`unknown log type: ${name}`)
-      break
+  if (name == 'console') {
+    os.print(console_rec(caller.line, caller.file, msg))
+  } else if (name == 'error') {
+    if (msg == null) msg = Error()
+    if (is_proto(msg, Error))
+      msg = msg.name + ": " + msg.message + "\n" + msg.stack
+    os.print(console_rec(caller.line, caller.file, msg))
+  } else if (name == 'system') {
+    msg = "[SYSTEM] " + msg
+    os.print(console_rec(caller.line, caller.file, msg))
+  } else {
+    log.console(`unknown log type: ${name}`)
   }
 }
 
-function disrupt(err)
+function actor_die(err)
 {
-  if (is_function(err.toString)) {
-  os.print(err.toString())
-  os.print("\n")
-  os.print(err.stack)
+  if (err && is_function(err.toString)) {
+    os.print(err.toString())
+    os.print("\n")
+    if (err.stack) os.print(err.stack)
   }
 
   if (overling) {
@@ -185,14 +186,14 @@ function disrupt(err)
       log.console(err.stack)
   }
 
-  actor_mod.disrupt()
+  actor_mod["disrupt"]()
 }
 
 
 
-actor_mod.on_exception(disrupt)
+actor_mod.on_exception(actor_die)
 
-_cell.args = init ?? {}
+_cell.args = init != null ? init : {}
 _cell.id = "newguy"
 
 function create_actor(desc = {id:guid()}) {
@@ -241,11 +242,15 @@ os.runtime_env = runtime_env
 
 $_.time_limit = function(requestor, seconds)
 {
-  if (!pronto.is_requestor(requestor))
-    throw Error('time_limit: first argument must be a requestor');
-  if (!is_number(seconds) || seconds <= 0)
-    throw Error('time_limit: seconds must be a positive number');
-    
+  if (!pronto.is_requestor(requestor)) {
+    log.error('time_limit: first argument must be a requestor')
+    disrupt
+  }
+  if (!is_number(seconds) || seconds <= 0) {
+    log.error('time_limit: seconds must be a positive number')
+    disrupt
+  }
+
   return function time_limit_requestor(callback, value) {
     pronto.check_callback(callback, 'time_limit')
     var finished = false
@@ -260,7 +265,14 @@ $_.time_limit = function(requestor, seconds)
         timer_cancel = null
       }
       if (requestor_cancel) {
-        try { pronto.requestor_cancel(reason) } catch (_) {}
+        requestor_cancel(reason)
+        requestor_cancel = null
+      }
+    }
+
+    function safe_cancel_requestor(reason) {
+      if (requestor_cancel) {
+        requestor_cancel(reason)
         requestor_cancel = null
       }
     }
@@ -268,15 +280,12 @@ $_.time_limit = function(requestor, seconds)
     timer_cancel = $_.delay(function() {
       if (finished) return
       def reason = make_reason(factory, 'Timeout.', seconds)
-      if (requestor_cancel) {
-        try { requestor_cancel(reason) } catch (_) {}
-        requestor_cancel = null
-      }
+      safe_cancel_requestor(reason)
       finished = true
       callback(null, reason)
     }, seconds)
 
-        try {
+    function do_request() {
       requestor_cancel = requestor(function(val, reason) {
         if (finished) return
         finished = true
@@ -286,16 +295,14 @@ $_.time_limit = function(requestor, seconds)
         }
         callback(val, reason)
       }, value)
-    } catch (ex) {
-      cancel(ex)
-      callback(null, ex)
+    } disruption {
+      cancel(Error('requestor failed'))
+      callback(null, Error('requestor failed'))
     }
+    do_request()
 
     return function(reason) {
-      if (requestor_cancel) {
-        try { requestor_cancel(reason) } catch (_) {}
-        requestor_cancel = null
-      }
+      safe_cancel_requestor(reason)
     }
   }
 }
@@ -408,52 +415,54 @@ var portal_fn = null
 
 // takes a function input value that will eventually be called with the current time in number form.
 $_.portal = function(fn, port) {
-  if (portal) throw Error(`Already started a portal listening on ${portal.port}`)
-  if (!port) throw Error("Requires a valid port.")
+  if (portal) {
+    log.error(`Already started a portal listening on ${portal.port}`)
+    disrupt
+  }
+  if (!port) {
+    log.error("Requires a valid port.")
+    disrupt
+  }
   log.system(`starting a portal on port ${port}`)
   portal = enet.create_host({address: "any", port})
   portal_fn = fn
 }
 
 function handle_host(e) {
-  switch (e.type) {
-    case "connect":
-      log.system(`connected a new peer: ${e.peer.address}:${e.peer.port}`)
-      peers[`${e.peer.address}:${e.peer.port}`] = e.peer
-      var queue = peer_queue.get(e.peer)
-      if (queue) {
-        arrfor(queue, (msg, index) => e.peer.send(nota.encode(msg)))
-        log.system(`sent ${msg} out of queue`)
-        peer_queue.delete(e.peer)
-      }
-      break
-    case "disconnect":
+  if (e.type == "connect") {
+    log.system(`connected a new peer: ${e.peer.address}:${e.peer.port}`)
+    peers[`${e.peer.address}:${e.peer.port}`] = e.peer
+    var queue = peer_queue.get(e.peer)
+    if (queue) {
+      arrfor(queue, (msg, index) => e.peer.send(nota.encode(msg)))
+      log.system(`sent ${msg} out of queue`)
       peer_queue.delete(e.peer)
-      arrfor(array(peers), function(id, index) {
-        if (peers[id] == e.peer) delete peers[id]
+    }
+  } else if (e.type == "disconnect") {
+    peer_queue.delete(e.peer)
+    arrfor(array(peers), function(id, index) {
+      if (peers[id] == e.peer) delete peers[id]
+    })
+    log.system('portal got disconnect from ' + e.peer.address + ":" + e.peer.port)
+  } else if (e.type == "receive") {
+    var data = nota.decode(e.data)
+    if (data.replycc && !data.replycc.address) {
+      data.replycc[ACTORDATA].address = e.peer.address
+      data.replycc[ACTORDATA].port = e.peer.port
+    }
+    function populate_actor_addresses(obj) {
+      if (!is_object(obj)) return
+      if (obj[ACTORDATA] && !obj[ACTORDATA].address) {
+        obj[ACTORDATA].address = e.peer.address
+        obj[ACTORDATA].port = e.peer.port
+      }
+      arrfor(array(obj), function(key, index) {
+        if (key in obj)
+          populate_actor_addresses(obj[key])
       })
-      log.system('portal got disconnect from ' + e.peer.address + ":" + e.peer.port)
-      break
-    case "receive":
-      var data = nota.decode(e.data)
-      if (data.replycc && !data.replycc.address) {
-        data.replycc[ACTORDATA].address = e.peer.address
-        data.replycc[ACTORDATA].port = e.peer.port
-      }
-      function populate_actor_addresses(obj) {
-        if (!is_object(obj)) return
-        if (obj[ACTORDATA] && !obj[ACTORDATA].address) {
-          obj[ACTORDATA].address = e.peer.address
-          obj[ACTORDATA].port = e.peer.port
-        }
-        arrfor(array(obj), function(key, index) {
-          if (key in obj)
-            populate_actor_addresses(obj[key])
-        })
-      }
-      if (data.data) populate_actor_addresses(data.data)
-      turn(data)
-      break
+    }
+    if (data.data) populate_actor_addresses(data.data)
+    turn(data)
   }
 }
 
@@ -487,10 +496,14 @@ $_.stop = function stop(actor) {
     need_stop = true
     return
   }
-  if (!is_actor(actor))
-    throw Error('Can only call stop on an actor.')
-  if (is_null(underlings[actor[ACTORDATA].id]))
-    throw Error('Can only call stop on an underling or self.')
+  if (!is_actor(actor)) {
+    log.error('Can only call stop on an actor.')
+    disrupt
+  }
+  if (is_null(underlings[actor[ACTORDATA].id])) {
+    log.error('Can only call stop on an underling or self.')
+    disrupt
+  }
 
   sys_msg(actor, {kind:"stop"})
 }
@@ -527,20 +540,22 @@ function actor_prep(actor, send) {
 
 // Send a message immediately without queuing
 function actor_send_immediate(actor, send) {
-  try {
-    actor_send(actor, send);
-  } catch (err) {
-    log.error("Failed to send immediate message:", err);
-  }
+  actor_send(actor, send)
 }
 
 function actor_send(actor, message) {
   if (actor[HEADER] && !actor[HEADER].replycc) // attempting to respond to a message but sender is not expecting; silently drop
     return
 
-  if (!is_actor(actor) && !is_actor(actor.replycc)) throw Error(`Must send to an actor object. Attempted send to ${actor}`)
-  
-  if (!is_object(message)) throw Error('Must send an object record.')
+  if (!is_actor(actor) && !is_actor(actor.replycc)) {
+    log.error(`Must send to an actor object. Attempted send to ${actor}`)
+    disrupt
+  }
+
+  if (!is_object(message)) {
+    log.error('Must send an object record.')
+    disrupt
+  }
 
   // message to self
   if (actor[ACTORDATA].id == _cell.id) {
@@ -583,12 +598,10 @@ function actor_send(actor, message) {
 // Holds all messages queued during the current turn.
 var message_queue = []
 
-var need_stop = false
-
- function send_messages() {
+function send_messages() {
    // if we've been flagged to stop, bail out before doing anything
    if (need_stop) {
-     disrupt()
+     actor_die()
      message_queue = []
      return
    }
@@ -608,19 +621,26 @@ var need_stop = false
 var replies = {}
 
 function send(actor, message, reply) {
-  if (!is_object(actor))
-    throw Error(`Must send to an actor object. Provided: ${actor}`);
+  if (!is_object(actor)) {
+    log.error(`Must send to an actor object. Provided: ${actor}`)
+    disrupt
+  }
 
-  if (!is_object(message))
-    throw Error('Message must be an object')
+  if (!is_object(message)) {
+    log.error('Message must be an object')
+    disrupt
+  }
   var send_msg = {type:"user", data: message}
+  var target = actor
 
   if (actor[HEADER] && actor[HEADER].replycc) {
     var header = actor[HEADER]
-    if (!header.replycc || !is_actor(header.replycc))
-      throw Error(`Supplied actor had a return, but it's not a valid actor! ${actor[HEADER]}`)
+    if (!header.replycc || !is_actor(header.replycc)) {
+      log.error(`Supplied actor had a return, but it's not a valid actor! ${actor[HEADER]}`)
+      disrupt
+    }
 
-    actor = header.replycc
+    target = header.replycc
     send_msg.return = header.reply
   }
 
@@ -638,7 +658,7 @@ function send(actor, message, reply) {
   }
 
   // Instead of sending immediately, queue it
-  actor_prep(actor, send_msg);
+  actor_prep(target, send_msg);
 }
 
 stone(send)
@@ -669,7 +689,7 @@ overling = _cell.args.overling
 $_.overling = overling
 
 root = _cell.args.root
-root ??= $_.self
+if (root == null) root = $_.self
 
 if (overling) {
   $_.couple(overling) // auto couple to overling
@@ -705,36 +725,35 @@ function handle_actor_disconnect(id) {
     delete greeters[id]
   }
   log.system(`actor ${id} disconnected`)
-  if (!is_null(couplings[id])) disrupt("coupled actor died") // couplings now disrupts instead of stop
+  if (!is_null(couplings[id])) actor_die("coupled actor died") // couplings now disrupts instead of stop
 }
 
 function handle_sysym(msg)
 {
   var from
-  switch(msg.kind) {
-    case 'stop':
-      disrupt("got stop message")
-      break
-    case 'underling':
-      from = msg.from
-      var greeter = greeters[from[ACTORDATA].id]
-      if (greeter) greeter(msg.message)
-      if (msg.message.type == 'disrupt')
-        delete underlings[from[ACTORDATA].id]
-      break
-    case 'contact':
-      if (portal_fn) {
-        var letter2 = msg.data
-        letter2[HEADER] = msg
-        delete msg.data
-        portal_fn(letter2)
-      } else throw Error('Got a contact message, but no portal is established.')
-      break
-    case 'couple': // from must be notified when we die
-      from = msg.from
-      underlings[from[ACTORDATA].id] = true
-      log.system(`actor ${from} is coupled to me`)
-      break
+  if (msg.kind == 'stop') {
+    actor_die("got stop message")
+  } else if (msg.kind == 'underling') {
+    from = msg.from
+    var greeter = greeters[from[ACTORDATA].id]
+    if (greeter) greeter(msg.message)
+    if (msg.message.type == 'disrupt')
+      delete underlings[from[ACTORDATA].id]
+  } else if (msg.kind == 'contact') {
+    if (portal_fn) {
+      var letter2 = msg.data
+      letter2[HEADER] = msg
+      delete msg.data
+      portal_fn(letter2)
+    } else {
+      log.error('Got a contact message, but no portal is established.')
+      disrupt
+    }
+  } else if (msg.kind == 'couple') {
+    // from must be notified when we die
+    from = msg.from
+    underlings[from[ACTORDATA].id] = true
+    log.system(`actor ${from} is coupled to me`)
   }
 }
 
@@ -744,30 +763,27 @@ function handle_message(msg) {
     return
   }
 
-  switch (msg.type) {
-    case "user":
-      var letter = msg.data // what the sender really sent
-      _ObjectDefineProperty(letter, HEADER, {
-        value: msg, enumerable: false
-      })
-      _ObjectDefineProperty(letter, ACTORDATA, { // this is so is_actor == true
-        value: { reply: msg.reply }, enumerable: false
-      })
-      
-      if (msg.return) {
-        var fn = replies[msg.return]
-        if (fn) fn(letter)
-        delete replies[msg.return]
-        return
-      }
-      
-      if (receive_fn) receive_fn(letter)
+  if (msg.type == "user") {
+    var letter = msg.data // what the sender really sent
+    _ObjectDefineProperty(letter, HEADER, {
+      value: msg, enumerable: false
+    })
+    _ObjectDefineProperty(letter, ACTORDATA, { // this is so is_actor == true
+      value: { reply: msg.reply }, enumerable: false
+    })
+
+    if (msg.return) {
+      var fn = replies[msg.return]
+      if (fn) fn(letter)
+      delete replies[msg.return]
       return
-    case "stopped":
-      handle_actor_disconnect(msg.id)
-      break
+    }
+
+    if (receive_fn) receive_fn(letter)
+  } else if (msg.type == "stopped") {
+    handle_actor_disconnect(msg.id)
   }
-};
+}
 
 function enet_check()
 {
@@ -792,8 +808,10 @@ if (!locator) {
   locator = shop.resolve_locator(_cell.args.program + ".ce", pkg)
 }
 
-if (!locator)
-  throw Error(`Main program ${_cell.args.program} could not be found`)
+if (!locator) {
+  os.print(`Main program ${_cell.args.program} could not be found\n`)
+  os.exit(1)
+}
 
 $_.clock(_ => {
   // Get capabilities for the main program
@@ -818,7 +836,6 @@ $_.clock(_ => {
   var val = call(locator.symbol, null, [_cell.args.arg, use_fn, env])
 
   if (val)
-    throw Error('Program must not return anything');
+    log.error('Program must not return anything')
+    disrupt
 })
-
-})()

internal/os.c

@@ -577,8 +577,8 @@ static const JSCFunctionListEntry js_os_funcs[] = {
 
 JSValue js_os_use(JSContext *js) {
   JS_NewClassID(&js_dylib_class_id);
-  JS_NewClass(JS_GetRuntime(js), js_dylib_class_id, &js_dylib_class);
-  
+  JS_NewClass(js, js_dylib_class_id, &js_dylib_class);
+
   JSValue mod = JS_NewObject(js);
   JS_SetPropertyFunctionList(js,mod,js_os_funcs,countof(js_os_funcs));
   return mod;

mcode.ce

@@ -0,0 +1,11 @@
+var fd = use("fd")
+var json = use("json")
+var tokenize = use("tokenize")
+var parse = use("parse")
+var mcode = use("mcode")
+var filename = args[0]
+var src = text(fd.slurp(filename))
+var result = tokenize(src, filename)
+var ast = parse(result.tokens, src, filename)
+var compiled = mcode(ast)
+print(json.encode(compiled))

meson.build

@@ -44,11 +44,14 @@ src += [        # core
   'wildmatch.c',
   'qjs_actor.c',
   'miniz.c',
-  'quickjs.c',
+  'runtime.c',
+  'mach.c',
+  'mcode.c',
   'libregexp.c', 'libunicode.c', 'cutils.c', 'dtoa.c'
 ]
 
 src += ['scheduler.c']
+src += ['qbe_helpers.c']
 
 scripts = [
   'debug/js.c',

net/enet.c

@@ -571,13 +571,13 @@ static const JSCFunctionListEntry js_enet_peer_funcs[] = {
 JSValue js_enet_use(JSContext *ctx)
 {
   JS_NewClassID(&enet_host_id);
-  JS_NewClass(JS_GetRuntime(ctx), enet_host_id, &enet_host);
+  JS_NewClass(ctx, enet_host_id, &enet_host);
   JSValue host_proto = JS_NewObject(ctx);
   JS_SetPropertyFunctionList(ctx, host_proto, js_enet_host_funcs, countof(js_enet_host_funcs));
   JS_SetClassProto(ctx, enet_host_id, host_proto);
 
   JS_NewClassID(&enet_peer_class_id);
-  JS_NewClass(JS_GetRuntime(ctx), enet_peer_class_id, &enet_peer_class);
+  JS_NewClass(ctx, enet_peer_class_id, &enet_peer_class);
   JSValue peer_proto = JS_NewObject(ctx);
   JS_SetPropertyFunctionList(ctx, peer_proto, js_enet_peer_funcs, countof(js_enet_peer_funcs));
   JS_SetClassProto(ctx, enet_peer_class_id, peer_proto);

parse.ce

@@ -0,0 +1,8 @@
+var fd = use("fd")
+var tokenize = use("tokenize")
+var parse = use("parse")
+var filename = args[0]
+var src = text(fd.slurp(filename))
+var result = tokenize(src, filename)
+var ast = parse(result.tokens, src, filename)
+print(json.encode(ast))

qbe.cm

@@ -0,0 +1,764 @@
+// QBE IL Macro Layer
+// Maps mcode VM operations to QBE intermediate language fragments.
+// Each macro returns a backtick template string of QBE IL.
+// Convention: p = unique prefix for temporaries/labels, result in %{p}
+
+// ============================================================
+// Constants
+// ============================================================
+
+def js_null = 7
+def js_false = 3
+def js_true = 35
+def js_exception = 15
+def js_empty_text = 27
+
+def int32_min = -2147483648
+def int32_max = 2147483647
+def mantissa_mask = 4503599627370495
+
+// ============================================================
+// Type Checks — each takes (p, v), produces %{p} as w (0 or 1)
+// ============================================================
+
+var is_int = function(p, v) {
+  return `  %${p}.t =l and ${v}, 1
+  %${p} =w ceql %${p}.t, 0
+`
+}
+
+var is_number = function(p, v) {
+  return `  %${p}.t1 =l and ${v}, 1
+  %${p}.ii =w ceql %${p}.t1, 0
+  %${p}.t2 =l and ${v}, 7
+  %${p}.fi =w ceql %${p}.t2, 5
+  %${p} =w or %${p}.ii, %${p}.fi
+`
+}
+
+var is_null = function(p, v) {
+  return `  %${p}.t =l and ${v}, 31
+  %${p} =w ceql %${p}.t, 7
+`
+}
+
+var is_bool = function(p, v) {
+  return `  %${p}.t =l and ${v}, 31
+  %${p} =w ceql %${p}.t, 3
+`
+}
+
+var is_exception = function(p, v) {
+  return `  %${p}.t =l and ${v}, 31
+  %${p} =w ceql %${p}.t, 15
+`
+}
+
+var is_ptr = function(p, v) {
+  return `  %${p}.t =l and ${v}, 7
+  %${p} =w ceql %${p}.t, 1
+`
+}
+
+var is_imm_text = function(p, v) {
+  return `  %${p}.t =l and ${v}, 31
+  %${p} =w ceql %${p}.t, 27
+`
+}
+
+var is_text = function(p, v) {
+  return `  %${p}.imm =l and ${v}, 31
+  %${p}.is_imm =w ceql %${p}.imm, 27
+  jnz %${p}.is_imm, @${p}.yes, @${p}.chk_ptr
+@${p}.chk_ptr
+  %${p}.ptag =l and ${v}, 7
+  %${p}.is_ptr =w ceql %${p}.ptag, 1
+  jnz %${p}.is_ptr, @${p}.load_hdr, @${p}.no
+@${p}.load_hdr
+  %${p}.ptr =l and ${v}, -8
+  %${p}.hdr =l loadl %${p}.ptr
+@${p}.chase
+  %${p}.ht =l and %${p}.hdr, 7
+  %${p}.is_fwd =w ceql %${p}.ht, 7
+  jnz %${p}.is_fwd, @${p}.follow, @${p}.chk_type
+@${p}.follow
+  %${p}.ptr =l shr %${p}.hdr, 3
+  %${p}.hdr =l loadl %${p}.ptr
+  jmp @${p}.chase
+@${p}.chk_type
+  %${p} =w ceql %${p}.ht, 2
+  jmp @${p}.done
+@${p}.yes
+  %${p} =w copy 1
+  jmp @${p}.done
+@${p}.no
+  %${p} =w copy 0
+@${p}.done
+`
+}
+
+// ============================================================
+// Value Extraction
+// ============================================================
+
+// get_int(p, v) — extract int32 as w from tagged int. Result: %{p}
+var get_int = function(p, v) {
+  return `  %${p}.sl =l sar ${v}, 1
+  %${p} =w copy %${p}.sl
+`
+}
+
+// get_bool(p, v) — extract bool as w. Result: %{p}
+var get_bool = function(p, v) {
+  return `  %${p}.sl =l shr ${v}, 5
+  %${p} =w and %${p}.sl, 1
+`
+}
+
+// get_ptr(p, v) — extract pointer as l. Result: %{p}
+var get_ptr = function(p, v) {
+  return `  %${p} =l and ${v}, -8
+`
+}
+
+// get_float64(p, v) — decode short float to d. Result: %{p} as d
+// Caller must handle zero check (sexp == 0 -> 0.0) before calling.
+var get_float64 = function(p, v) {
+  return `  %${p}.sign =l shr ${v}, 63
+  %${p}.sexp =l shr ${v}, 55
+  %${p}.sexp =l and %${p}.sexp, 255
+  %${p}.mant =l shr ${v}, 3
+  %${p}.mant =l and %${p}.mant, ${mantissa_mask}
+  %${p}.dexp =l sub %${p}.sexp, 127
+  %${p}.dexp =l add %${p}.dexp, 1023
+  %${p}.s63 =l shl %${p}.sign, 63
+  %${p}.e52 =l shl %${p}.dexp, 52
+  %${p}.bits =l or %${p}.s63, %${p}.e52
+  %${p}.bits =l or %${p}.bits, %${p}.mant
+  %${p} =d cast %${p}.bits
+`
+}
+
+// to_float64(p, v) — convert any numeric value (int or short float) to d.
+// Result: %{p} as d
+var to_float64 = function(p, v) {
+  return `  %${p}.tag =l and ${v}, 1
+  %${p}.is_int =w ceql %${p}.tag, 0
+  jnz %${p}.is_int, @${p}.from_int, @${p}.from_float
+@${p}.from_int
+  %${p}.isl =l sar ${v}, 1
+  %${p}.iw =w copy %${p}.isl
+  %${p} =d swtof %${p}.iw
+  jmp @${p}.done
+@${p}.from_float
+  %${p}.fsexp =l shr ${v}, 55
+  %${p}.fsexp =l and %${p}.fsexp, 255
+  %${p}.is_zero =w ceql %${p}.fsexp, 0
+  jnz %${p}.is_zero, @${p}.fzero, @${p}.fdecode
+@${p}.fzero
+  %${p} =d copy d_0.0
+  jmp @${p}.done
+@${p}.fdecode
+  %${p}.fsign =l shr ${v}, 63
+  %${p}.fmant =l shr ${v}, 3
+  %${p}.fmant =l and %${p}.fmant, ${mantissa_mask}
+  %${p}.fdexp =l sub %${p}.fsexp, 127
+  %${p}.fdexp =l add %${p}.fdexp, 1023
+  %${p}.fs63 =l shl %${p}.fsign, 63
+  %${p}.fe52 =l shl %${p}.fdexp, 52
+  %${p}.fbits =l or %${p}.fs63, %${p}.fe52
+  %${p}.fbits =l or %${p}.fbits, %${p}.fmant
+  %${p} =d cast %${p}.fbits
+@${p}.done
+`
+}
+
+// ============================================================
+// Value Creation
+// ============================================================
+
+// new_int(p, i) — tag int32 w as JSValue l. Result: %{p}
+var new_int = function(p, i) {
+  return `  %${p}.ext =l extuw ${i}
+  %${p} =l shl %${p}.ext, 1
+`
+}
+
+// new_bool(p, b) — tag bool w as JSValue l. Result: %{p}
+var new_bool = function(p, b) {
+  return `  %${p}.ext =l extuw ${b}
+  %${p}.sh =l shl %${p}.ext, 5
+  %${p} =l or %${p}.sh, 3
+`
+}
+
+// new_float64 — C call to __JS_NewFloat64(ctx, val). Result: %{p}
+var new_float64 = function(p, ctx, d) {
+  return `  %${p} =l call $__JS_NewFloat64(l ${ctx}, d ${d})
+`
+}
+
+// ============================================================
+// Arithmetic — add(p, ctx, a, b)
+// Int fast path inline, text concat and float as C calls.
+// Jumps to @disrupt on type mismatch.
+// ============================================================
+
+var add = function(p, ctx, a, b) {
+  return `@${p}.start
+  %${p}.at =l and ${a}, 1
+  %${p}.bt =l and ${b}, 1
+  %${p}.not_int =l or %${p}.at, %${p}.bt
+  jnz %${p}.not_int, @${p}.not_both_int, @${p}.int_path
+@${p}.int_path
+  %${p}.ia =l sar ${a}, 1
+  %${p}.ib =l sar ${b}, 1
+  %${p}.sum =l add %${p}.ia, %${p}.ib
+  %${p}.lo =w csltl %${p}.sum, ${int32_min}
+  %${p}.hi =w csgtl %${p}.sum, ${int32_max}
+  %${p}.ov =w or %${p}.lo, %${p}.hi
+  jnz %${p}.ov, @${p}.int_overflow, @${p}.int_ok
+@${p}.int_ok
+  %${p}.rw =w copy %${p}.sum
+  %${p}.rext =l extuw %${p}.rw
+  %${p} =l shl %${p}.rext, 1
+  jmp @${p}.done
+@${p}.int_overflow
+  %${p}.fd =d sltof %${p}.sum
+  %${p} =l call $__JS_NewFloat64(l ${ctx}, d %${p}.fd)
+  jmp @${p}.done
+@${p}.not_both_int
+  %${p}.a_is_text =w call $JS_IsText(l ${a})
+  %${p}.b_is_text =w call $JS_IsText(l ${b})
+  %${p}.both_text =w and %${p}.a_is_text, %${p}.b_is_text
+  jnz %${p}.both_text, @${p}.text_path, @${p}.chk_num
+@${p}.text_path
+  %${p} =l call $JS_ConcatString(l ${ctx}, l ${a}, l ${b})
+  jmp @${p}.done
+@${p}.chk_num
+  %${p}.a_is_num =w call $JS_IsNumber(l ${a})
+  %${p}.b_is_num =w call $JS_IsNumber(l ${b})
+  %${p}.both_num =w and %${p}.a_is_num, %${p}.b_is_num
+  jnz %${p}.both_num, @${p}.float_path, @disrupt
+@${p}.float_path
+  %${p} =l call $qbe_float_add(l ${ctx}, l ${a}, l ${b})
+@${p}.done
+`
+}
+
+var sub = function(p, ctx, a, b) {
+  return `@${p}.start
+  %${p}.at =l and ${a}, 1
+  %${p}.bt =l and ${b}, 1
+  %${p}.not_int =l or %${p}.at, %${p}.bt
+  jnz %${p}.not_int, @${p}.not_both_int, @${p}.int_path
+@${p}.int_path
+  %${p}.ia =l sar ${a}, 1
+  %${p}.ib =l sar ${b}, 1
+  %${p}.diff =l sub %${p}.ia, %${p}.ib
+  %${p}.lo =w csltl %${p}.diff, ${int32_min}
+  %${p}.hi =w csgtl %${p}.diff, ${int32_max}
+  %${p}.ov =w or %${p}.lo, %${p}.hi
+  jnz %${p}.ov, @${p}.int_overflow, @${p}.int_ok
+@${p}.int_ok
+  %${p}.rw =w copy %${p}.diff
+  %${p}.rext =l extuw %${p}.rw
+  %${p} =l shl %${p}.rext, 1
+  jmp @${p}.done
+@${p}.int_overflow
+  %${p}.fd =d sltof %${p}.diff
+  %${p} =l call $__JS_NewFloat64(l ${ctx}, d %${p}.fd)
+  jmp @${p}.done
+@${p}.not_both_int
+  %${p}.a_is_num =w call $JS_IsNumber(l ${a})
+  %${p}.b_is_num =w call $JS_IsNumber(l ${b})
+  %${p}.both_num =w and %${p}.a_is_num, %${p}.b_is_num
+  jnz %${p}.both_num, @${p}.float_path, @disrupt
+@${p}.float_path
+  %${p} =l call $qbe_float_sub(l ${ctx}, l ${a}, l ${b})
+@${p}.done
+`
+}
+
+var mul = function(p, ctx, a, b) {
+  return `@${p}.start
+  %${p}.at =l and ${a}, 1
+  %${p}.bt =l and ${b}, 1
+  %${p}.not_int =l or %${p}.at, %${p}.bt
+  jnz %${p}.not_int, @${p}.not_both_int, @${p}.int_path
+@${p}.int_path
+  %${p}.ia =l sar ${a}, 1
+  %${p}.ib =l sar ${b}, 1
+  %${p}.prod =l mul %${p}.ia, %${p}.ib
+  %${p}.lo =w csltl %${p}.prod, ${int32_min}
+  %${p}.hi =w csgtl %${p}.prod, ${int32_max}
+  %${p}.ov =w or %${p}.lo, %${p}.hi
+  jnz %${p}.ov, @${p}.int_overflow, @${p}.int_ok
+@${p}.int_ok
+  %${p}.rw =w copy %${p}.prod
+  %${p}.rext =l extuw %${p}.rw
+  %${p} =l shl %${p}.rext, 1
+  jmp @${p}.done
+@${p}.int_overflow
+  %${p}.fd =d sltof %${p}.prod
+  %${p} =l call $__JS_NewFloat64(l ${ctx}, d %${p}.fd)
+  jmp @${p}.done
+@${p}.not_both_int
+  %${p}.a_is_num =w call $JS_IsNumber(l ${a})
+  %${p}.b_is_num =w call $JS_IsNumber(l ${b})
+  %${p}.both_num =w and %${p}.a_is_num, %${p}.b_is_num
+  jnz %${p}.both_num, @${p}.float_path, @disrupt
+@${p}.float_path
+  %${p} =l call $qbe_float_mul(l ${ctx}, l ${a}, l ${b})
+@${p}.done
+`
+}
+
+var div = function(p, ctx, a, b) {
+  return `@${p}.start
+  %${p}.at =l and ${a}, 1
+  %${p}.bt =l and ${b}, 1
+  %${p}.not_int =l or %${p}.at, %${p}.bt
+  jnz %${p}.not_int, @${p}.not_both_int, @${p}.int_path
+@${p}.int_path
+  %${p}.ia =w copy 0
+  %${p}.tmp =l sar ${a}, 1
+  %${p}.ia =w copy %${p}.tmp
+  %${p}.ib =w copy 0
+  %${p}.tmp2 =l sar ${b}, 1
+  %${p}.ib =w copy %${p}.tmp2
+  %${p}.div0 =w ceqw %${p}.ib, 0
+  jnz %${p}.div0, @${p}.ret_null, @${p}.chk_exact
+@${p}.ret_null
+  %${p} =l copy ${js_null}
+  jmp @${p}.done
+@${p}.chk_exact
+  %${p}.rem =w rem %${p}.ia, %${p}.ib
+  %${p}.exact =w ceqw %${p}.rem, 0
+  jnz %${p}.exact, @${p}.int_div, @${p}.int_to_float
+@${p}.int_div
+  %${p}.q =w div %${p}.ia, %${p}.ib
+  %${p}.qext =l extuw %${p}.q
+  %${p} =l shl %${p}.qext, 1
+  jmp @${p}.done
+@${p}.int_to_float
+  %${p}.da =d swtof %${p}.ia
+  %${p}.db =d swtof %${p}.ib
+  %${p}.dr =d div %${p}.da, %${p}.db
+  %${p} =l call $__JS_NewFloat64(l ${ctx}, d %${p}.dr)
+  jmp @${p}.done
+@${p}.not_both_int
+  %${p}.a_is_num =w call $JS_IsNumber(l ${a})
+  %${p}.b_is_num =w call $JS_IsNumber(l ${b})
+  %${p}.both_num =w and %${p}.a_is_num, %${p}.b_is_num
+  jnz %${p}.both_num, @${p}.float_path, @disrupt
+@${p}.float_path
+  %${p} =l call $qbe_float_div(l ${ctx}, l ${a}, l ${b})
+@${p}.done
+`
+}
+
+var mod = function(p, ctx, a, b) {
+  return `@${p}.start
+  %${p}.at =l and ${a}, 1
+  %${p}.bt =l and ${b}, 1
+  %${p}.not_int =l or %${p}.at, %${p}.bt
+  jnz %${p}.not_int, @${p}.not_both_int, @${p}.int_path
+@${p}.int_path
+  %${p}.ia =w copy 0
+  %${p}.tmp =l sar ${a}, 1
+  %${p}.ia =w copy %${p}.tmp
+  %${p}.ib =w copy 0
+  %${p}.tmp2 =l sar ${b}, 1
+  %${p}.ib =w copy %${p}.tmp2
+  %${p}.div0 =w ceqw %${p}.ib, 0
+  jnz %${p}.div0, @${p}.ret_null, @${p}.do_mod
+@${p}.ret_null
+  %${p} =l copy ${js_null}
+  jmp @${p}.done
+@${p}.do_mod
+  %${p}.r =w rem %${p}.ia, %${p}.ib
+  %${p}.rext =l extuw %${p}.r
+  %${p} =l shl %${p}.rext, 1
+  jmp @${p}.done
+@${p}.not_both_int
+  %${p}.a_is_num =w call $JS_IsNumber(l ${a})
+  %${p}.b_is_num =w call $JS_IsNumber(l ${b})
+  %${p}.both_num =w and %${p}.a_is_num, %${p}.b_is_num
+  jnz %${p}.both_num, @${p}.float_path, @disrupt
+@${p}.float_path
+  %${p} =l call $qbe_float_mod(l ${ctx}, l ${a}, l ${b})
+@${p}.done
+`
+}
+
+// ============================================================
+// Comparisons — eq, ne, lt, le, gt, ge
+// Each takes (p, ctx, a, b), produces %{p} as l (tagged JSValue bool)
+// ============================================================
+
+// Helper: generate comparison for a given op string and int comparison QBE op
+// null_true: whether null==null returns true (eq, le, ge) or false (ne, lt, gt)
+var cmp = function(p, ctx, a, b, int_cmp_op, float_cmp_op_id, is_eq, is_ne, null_true) {
+  var eq_only = 0
+  if (is_eq || is_ne) {
+    eq_only = 1
+  }
+  var mismatch_val = js_false
+  if (is_ne) {
+    mismatch_val = js_true
+  }
+  var null_val = js_false
+  if (null_true) {
+    null_val = js_true
+  }
+  return `@${p}.start
+  %${p}.at =l and ${a}, 1
+  %${p}.bt =l and ${b}, 1
+  %${p}.not_int =l or %${p}.at, %${p}.bt
+  jnz %${p}.not_int, @${p}.not_both_int, @${p}.int_path
+@${p}.int_path
+  %${p}.ia =l sar ${a}, 1
+  %${p}.ib =l sar ${b}, 1
+  %${p}.iw =w copy %${p}.ia
+  %${p}.ibw =w copy %${p}.ib
+  %${p}.cr =w ${int_cmp_op} %${p}.iw, %${p}.ibw
+  %${p}.crext =l extuw %${p}.cr
+  %${p}.sh =l shl %${p}.crext, 5
+  %${p} =l or %${p}.sh, 3
+  jmp @${p}.done
+@${p}.not_both_int
+  %${p}.a_tag5 =l and ${a}, 31
+  %${p}.b_tag5 =l and ${b}, 31
+  %${p}.a_is_null =w ceql %${p}.a_tag5, 7
+  %${p}.b_is_null =w ceql %${p}.b_tag5, 7
+  %${p}.both_null =w and %${p}.a_is_null, %${p}.b_is_null
+  jnz %${p}.both_null, @${p}.null_path, @${p}.chk_bool
+@${p}.null_path
+  %${p} =l copy ${null_val}
+  jmp @${p}.done
+@${p}.chk_bool
+  %${p}.a_is_bool =w ceql %${p}.a_tag5, 3
+  %${p}.b_is_bool =w ceql %${p}.b_tag5, 3
+  %${p}.both_bool =w and %${p}.a_is_bool, %${p}.b_is_bool
+  jnz %${p}.both_bool, @${p}.bool_path, @${p}.chk_num
+@${p}.bool_path
+  %${p}.ba =l shr ${a}, 5
+  %${p}.baw =w and %${p}.ba, 1
+  %${p}.bb =l shr ${b}, 5
+  %${p}.bbw =w and %${p}.bb, 1
+  %${p}.bcr =w ${int_cmp_op} %${p}.baw, %${p}.bbw
+  %${p}.bcrext =l extuw %${p}.bcr
+  %${p}.bsh =l shl %${p}.bcrext, 5
+  %${p} =l or %${p}.bsh, 3
+  jmp @${p}.done
+@${p}.chk_num
+  %${p}.a_is_num =w call $JS_IsNumber(l ${a})
+  %${p}.b_is_num =w call $JS_IsNumber(l ${b})
+  %${p}.both_num =w and %${p}.a_is_num, %${p}.b_is_num
+  jnz %${p}.both_num, @${p}.num_path, @${p}.chk_text
+@${p}.num_path
+  %${p}.fcr =w call $qbe_float_cmp(l ${ctx}, w ${float_cmp_op_id}, l ${a}, l ${b})
+  %${p}.fcrext =l extuw %${p}.fcr
+  %${p}.fsh =l shl %${p}.fcrext, 5
+  %${p} =l or %${p}.fsh, 3
+  jmp @${p}.done
+@${p}.chk_text
+  %${p}.a_is_text =w call $JS_IsText(l ${a})
+  %${p}.b_is_text =w call $JS_IsText(l ${b})
+  %${p}.both_text =w and %${p}.a_is_text, %${p}.b_is_text
+  jnz %${p}.both_text, @${p}.text_path, @${p}.mismatch
+@${p}.text_path
+  %${p}.scmp =w call $js_string_compare_value(l ${ctx}, l ${a}, l ${b}, w ${eq_only})
+  %${p}.tcr =w ${int_cmp_op} %${p}.scmp, 0
+  %${p}.tcrext =l extuw %${p}.tcr
+  %${p}.tsh =l shl %${p}.tcrext, 5
+  %${p} =l or %${p}.tsh, 3
+  jmp @${p}.done
+@${p}.mismatch
+  %${p} =l copy ${mismatch_val}
+@${p}.done
+`
+}
+
+// Comparison op IDs matching MACH_EQ..MACH_GE order for qbe_float_cmp
+// MACH_EQ=0, NEQ=1, LT=2, LE=3, GT=4, GE=5
+// null_true: eq, le, ge return true for null==null; ne, lt, gt return false
+var eq = function(p, ctx, a, b) {
+  return cmp(p, ctx, a, b, "ceqw", 0, true, false, true)
+}
+
+var ne = function(p, ctx, a, b) {
+  return cmp(p, ctx, a, b, "cnew", 1, false, true, false)
+}
+
+var lt = function(p, ctx, a, b) {
+  return cmp(p, ctx, a, b, "csltw", 2, false, false, false)
+}
+
+var le = function(p, ctx, a, b) {
+  return cmp(p, ctx, a, b, "cslew", 3, false, false, true)
+}
+
+var gt = function(p, ctx, a, b) {
+  return cmp(p, ctx, a, b, "csgtw", 4, false, false, false)
+}
+
+var ge = function(p, ctx, a, b) {
+  return cmp(p, ctx, a, b, "csgew", 5, false, false, true)
+}
+
+// ============================================================
+// Unary Ops
+// ============================================================
+
+// neg(p, ctx, v) — negate. Int fast path (INT32_MIN edge case), else C call.
+var neg = function(p, ctx, v) {
+  return `@${p}.start
+  %${p}.tag =l and ${v}, 1
+  %${p}.is_int =w ceql %${p}.tag, 0
+  jnz %${p}.is_int, @${p}.int_path, @${p}.float_path
+@${p}.int_path
+  %${p}.sl =l sar ${v}, 1
+  %${p}.iw =w copy %${p}.sl
+  %${p}.is_min =w ceqw %${p}.iw, ${int32_min}
+  jnz %${p}.is_min, @${p}.min_overflow, @${p}.int_ok
+@${p}.min_overflow
+  %${p}.fd =d swtof %${p}.iw
+  %${p}.fdn =d neg %${p}.fd
+  %${p} =l call $__JS_NewFloat64(l ${ctx}, d %${p}.fdn)
+  jmp @${p}.done
+@${p}.int_ok
+  %${p}.ni =w sub 0, %${p}.iw
+  %${p}.niext =l extuw %${p}.ni
+  %${p} =l shl %${p}.niext, 1
+  jmp @${p}.done
+@${p}.float_path
+  %${p} =l call $qbe_float_neg(l ${ctx}, l ${v})
+@${p}.done
+`
+}
+
+// inc(p, ctx, v) — increment. Int fast path (INT32_MAX edge case), else C call.
+var inc = function(p, ctx, v) {
+  return `@${p}.start
+  %${p}.tag =l and ${v}, 1
+  %${p}.is_int =w ceql %${p}.tag, 0
+  jnz %${p}.is_int, @${p}.int_path, @${p}.float_path
+@${p}.int_path
+  %${p}.sl =l sar ${v}, 1
+  %${p}.iw =w copy %${p}.sl
+  %${p}.is_max =w ceqw %${p}.iw, ${int32_max}
+  jnz %${p}.is_max, @${p}.max_overflow, @${p}.int_ok
+@${p}.max_overflow
+  %${p}.fd =d swtof %${p}.iw
+  %${p}.fd1 =d add %${p}.fd, d_1.0
+  %${p} =l call $__JS_NewFloat64(l ${ctx}, d %${p}.fd1)
+  jmp @${p}.done
+@${p}.int_ok
+  %${p}.ni =w add %${p}.iw, 1
+  %${p}.niext =l extuw %${p}.ni
+  %${p} =l shl %${p}.niext, 1
+  jmp @${p}.done
+@${p}.float_path
+  %${p} =l call $qbe_float_inc(l ${ctx}, l ${v})
+@${p}.done
+`
+}
+
+// dec(p, ctx, v) — decrement. Int fast path (INT32_MIN edge case), else C call.
+var dec = function(p, ctx, v) {
+  return `@${p}.start
+  %${p}.tag =l and ${v}, 1
+  %${p}.is_int =w ceql %${p}.tag, 0
+  jnz %${p}.is_int, @${p}.int_path, @${p}.float_path
+@${p}.int_path
+  %${p}.sl =l sar ${v}, 1
+  %${p}.iw =w copy %${p}.sl
+  %${p}.is_min =w ceqw %${p}.iw, ${int32_min}
+  jnz %${p}.is_min, @${p}.min_overflow, @${p}.int_ok
+@${p}.min_overflow
+  %${p}.fd =d swtof %${p}.iw
+  %${p}.fd1 =d sub %${p}.fd, d_1.0
+  %${p} =l call $__JS_NewFloat64(l ${ctx}, d %${p}.fd1)
+  jmp @${p}.done
+@${p}.int_ok
+  %${p}.ni =w sub %${p}.iw, 1
+  %${p}.niext =l extuw %${p}.ni
+  %${p} =l shl %${p}.niext, 1
+  jmp @${p}.done
+@${p}.float_path
+  %${p} =l call $qbe_float_dec(l ${ctx}, l ${v})
+@${p}.done
+`
+}
+
+// lnot(p, ctx, v) — logical not. C call to JS_ToBool, then negate inline.
+var lnot = function(p, ctx, v) {
+  return `  %${p}.bval =w call $JS_ToBool(l ${ctx}, l ${v})
+  %${p}.neg =w ceqw %${p}.bval, 0
+  %${p}.nex =l extuw %${p}.neg
+  %${p}.sh =l shl %${p}.nex, 5
+  %${p} =l or %${p}.sh, 3
+`
+}
+
+// bnot(p, ctx, v) — bitwise not. Convert to int32, ~, re-tag.
+var bnot = function(p, ctx, v) {
+  return `@${p}.start
+  %${p}.tag =l and ${v}, 1
+  %${p}.is_int =w ceql %${p}.tag, 0
+  jnz %${p}.is_int, @${p}.int_path, @${p}.slow_path
+@${p}.int_path
+  %${p}.sl =l sar ${v}, 1
+  %${p}.iw =w copy %${p}.sl
+  %${p}.nw =w xor %${p}.iw, -1
+  %${p}.nex =l extuw %${p}.nw
+  %${p} =l shl %${p}.nex, 1
+  jmp @${p}.done
+@${p}.slow_path
+  %${p} =l call $qbe_bnot(l ${ctx}, l ${v})
+@${p}.done
+`
+}
+
+// ============================================================
+// Bitwise Ops — band, bor, bxor, shl, shr, ushr
+// Both operands must be numeric. Int fast path, float -> convert to int32.
+// ============================================================
+
+var bitwise_op = function(p, ctx, a, b, qbe_op) {
+  return `@${p}.start
+  %${p}.at =l and ${a}, 1
+  %${p}.bt =l and ${b}, 1
+  %${p}.not_int =l or %${p}.at, %${p}.bt
+  jnz %${p}.not_int, @${p}.slow_path, @${p}.int_path
+@${p}.int_path
+  %${p}.ia =l sar ${a}, 1
+  %${p}.iaw =w copy %${p}.ia
+  %${p}.ib =l sar ${b}, 1
+  %${p}.ibw =w copy %${p}.ib
+  %${p}.rw =w ${qbe_op} %${p}.iaw, %${p}.ibw
+  %${p}.rext =l extuw %${p}.rw
+  %${p} =l shl %${p}.rext, 1
+  jmp @${p}.done
+@${p}.slow_path
+  %${p}.a_is_num =w call $JS_IsNumber(l ${a})
+  %${p}.b_is_num =w call $JS_IsNumber(l ${b})
+  %${p}.both_num =w and %${p}.a_is_num, %${p}.b_is_num
+  jnz %${p}.both_num, @${p}.float_to_int, @disrupt
+@${p}.float_to_int
+  %${p} =l call $qbe_bitwise_${qbe_op}(l ${ctx}, l ${a}, l ${b})
+@${p}.done
+`
+}
+
+var band = function(p, ctx, a, b) {
+  return bitwise_op(p, ctx, a, b, "and")
+}
+
+var bor = function(p, ctx, a, b) {
+  return bitwise_op(p, ctx, a, b, "or")
+}
+
+var bxor = function(p, ctx, a, b) {
+  return bitwise_op(p, ctx, a, b, "xor")
+}
+
+// Shift ops: mask shift amount to 5 bits (& 31)
+var shift_op = function(p, ctx, a, b, qbe_op) {
+  return `@${p}.start
+  %${p}.at =l and ${a}, 1
+  %${p}.bt =l and ${b}, 1
+  %${p}.not_int =l or %${p}.at, %${p}.bt
+  jnz %${p}.not_int, @${p}.slow_path, @${p}.int_path
+@${p}.int_path
+  %${p}.ia =l sar ${a}, 1
+  %${p}.iaw =w copy %${p}.ia
+  %${p}.ib =l sar ${b}, 1
+  %${p}.ibw =w copy %${p}.ib
+  %${p}.sh =w and %${p}.ibw, 31
+  %${p}.rw =w ${qbe_op} %${p}.iaw, %${p}.sh
+  %${p}.rext =l extuw %${p}.rw
+  %${p} =l shl %${p}.rext, 1
+  jmp @${p}.done
+@${p}.slow_path
+  %${p}.a_is_num =w call $JS_IsNumber(l ${a})
+  %${p}.b_is_num =w call $JS_IsNumber(l ${b})
+  %${p}.both_num =w and %${p}.a_is_num, %${p}.b_is_num
+  jnz %${p}.both_num, @${p}.float_to_int, @disrupt
+@${p}.float_to_int
+  %${p} =l call $qbe_shift_${qbe_op}(l ${ctx}, l ${a}, l ${b})
+@${p}.done
+`
+}
+
+var shl = function(p, ctx, a, b) {
+  return shift_op(p, ctx, a, b, "shl")
+}
+
+var shr = function(p, ctx, a, b) {
+  return shift_op(p, ctx, a, b, "sar")
+}
+
+var ushr = function(p, ctx, a, b) {
+  return shift_op(p, ctx, a, b, "shr")
+}
+
+// ============================================================
+// Module export
+// ============================================================
+
+return {
+  // constants
+  js_null: js_null,
+  js_false: js_false,
+  js_true: js_true,
+  js_exception: js_exception,
+  js_empty_text: js_empty_text,
+  // type checks
+  is_int: is_int,
+  is_number: is_number,
+  is_null: is_null,
+  is_bool: is_bool,
+  is_exception: is_exception,
+  is_ptr: is_ptr,
+  is_imm_text: is_imm_text,
+  is_text: is_text,
+  // value extraction
+  get_int: get_int,
+  get_bool: get_bool,
+  get_ptr: get_ptr,
+  get_float64: get_float64,
+  to_float64: to_float64,
+  // value creation
+  new_int: new_int,
+  new_bool: new_bool,
+  new_float64: new_float64,
+  // arithmetic
+  add: add,
+  sub: sub,
+  mul: mul,
+  div: div,
+  mod: mod,
+  // comparisons
+  eq: eq,
+  ne: ne,
+  lt: lt,
+  le: le,
+  gt: gt,
+  ge: ge,
+  // unary
+  neg: neg,
+  inc: inc,
+  dec: dec,
+  lnot: lnot,
+  bnot: bnot,
+  // bitwise
+  band: band,
+  bor: bor,
+  bxor: bxor,
+  shl: shl,
+  shr: shr,
+  ushr: ushr
+}

qop.c

@@ -458,13 +458,13 @@ static const JSCFunctionListEntry js_qop_funcs[] = {
 
 JSValue js_qop_use(JSContext *js) {
     JS_NewClassID(&js_qop_archive_class_id);
-    JS_NewClass(JS_GetRuntime(js), js_qop_archive_class_id, &js_qop_archive_class);
+    JS_NewClass(js, js_qop_archive_class_id, &js_qop_archive_class);
     JSValue archive_proto = JS_NewObject(js);
     JS_SetPropertyFunctionList(js, archive_proto, js_qop_archive_funcs, countof(js_qop_archive_funcs));
     JS_SetClassProto(js, js_qop_archive_class_id, archive_proto);
 
     JS_NewClassID(&js_qop_writer_class_id);
-    JS_NewClass(JS_GetRuntime(js), js_qop_writer_class_id, &js_qop_writer_class);
+    JS_NewClass(js, js_qop_writer_class_id, &js_qop_writer_class);
     JSValue writer_proto = JS_NewObject(js);
     JS_SetPropertyFunctionList(js, writer_proto, js_qop_writer_funcs, countof(js_qop_writer_funcs));
     JS_SetClassProto(js, js_qop_writer_class_id, writer_proto);

source/cell.c

@@ -11,7 +11,8 @@
 #include "cell_internal.h"
 #include "cJSON.h"
 
-#define ENGINE "internal/engine.cm"
+#define BOOTSTRAP_MACH "internal/bootstrap.mach"
+#define BOOTSTRAP_AST  "internal/bootstrap.ast.json"
 #define CELL_SHOP_DIR ".cell"
 #define CELL_CORE_DIR "packages/core"
 
@@ -26,6 +27,7 @@ static int run_test_suite(size_t heap_size);
 
 cell_rt *root_cell = NULL;
 static char *core_path = NULL;
+static JSRuntime *g_runtime = NULL;
 
 // Get the home directory
 static const char* get_home_dir(void) {
@@ -70,70 +72,41 @@ int find_cell_shop(void)
 // Load a file from the core directory
 static char* load_core_file(const char *filename, size_t *out_size) {
   if (!core_path) return NULL;
-  
+
   size_t path_len = strlen(core_path) + 1 + strlen(filename) + 1;
   char *full_path = malloc(path_len);
   if (!full_path) return NULL;
-  
+
   snprintf(full_path, path_len, "%s/%s", core_path, filename);
-  
+
   FILE *fh = fopen(full_path, "rb");
   free(full_path);
-  
+
   if (!fh) return NULL;
-  
+
   fseek(fh, 0, SEEK_END);
   long file_size = ftell(fh);
   fseek(fh, 0, SEEK_SET);
-  
+
   char *data = malloc(file_size + 1);
   if (!data) {
     fclose(fh);
     return NULL;
   }
-  
+
   if (fread(data, 1, file_size, fh) != (size_t)file_size) {
     free(data);
     fclose(fh);
     return NULL;
   }
-  
+
   fclose(fh);
   data[file_size] = 0;
-  
+
   if (out_size) *out_size = file_size;
   return data;
 }
 
-static int print_json_errors(const char *json) {
-  if (!json) return 0;
-  cJSON *root = cJSON_Parse(json);
-  if (!root) return 0;
-  cJSON *errors = cJSON_GetObjectItemCaseSensitive(root, "errors");
-  if (!cJSON_IsArray(errors) || cJSON_GetArraySize(errors) == 0) {
-    cJSON_Delete(root);
-    return 0;
-  }
-  const char *filename = "<unknown>";
-  cJSON *fname = cJSON_GetObjectItemCaseSensitive(root, "filename");
-  if (cJSON_IsString(fname))
-    filename = fname->valuestring;
-  cJSON *e;
-  cJSON_ArrayForEach(e, errors) {
-    const char *msg = cJSON_GetStringValue(
-      cJSON_GetObjectItemCaseSensitive(e, "message"));
-    cJSON *line = cJSON_GetObjectItemCaseSensitive(e, "line");
-    cJSON *col = cJSON_GetObjectItemCaseSensitive(e, "column");
-    if (msg && cJSON_IsNumber(line) && cJSON_IsNumber(col))
-      fprintf(stderr, "%s:%d:%d: error: %s\n",
-        filename, (int)line->valuedouble, (int)col->valuedouble, msg);
-    else if (msg)
-      fprintf(stderr, "%s: error: %s\n", filename, msg);
-  }
-  cJSON_Delete(root);
-  return 1;
-}
-
 // Get the core path for use by scripts
 const char* cell_get_core_path(void) {
   return core_path;
@@ -154,28 +127,40 @@ JSValue js_wota_use(JSContext *js);
 
 void script_startup(cell_rt *prt)
 {
-  JSRuntime *rt = JS_NewRuntime();
-  JS_SetInterruptHandler(rt, (JSInterruptHandler *)actor_interrupt_cb, prt);
-  JSContext *js = JS_NewContext(rt);
+  if (!g_runtime) {
+    g_runtime = JS_NewRuntime();
+  }
+  JSContext *js = JS_NewContext(g_runtime);
+  JS_SetInterruptHandler(js, (JSInterruptHandler *)actor_interrupt_cb, prt);
 
   JS_SetContextOpaque(js, prt);
   prt->context = js;
 
+  /* Register all GCRef fields so the Cheney GC can relocate them. */
+  JS_AddGCRef(js, &prt->idx_buffer_ref);
+  JS_AddGCRef(js, &prt->on_exception_ref);
+  JS_AddGCRef(js, &prt->message_handle_ref);
+  JS_AddGCRef(js, &prt->unneeded_ref);
+  JS_AddGCRef(js, &prt->actor_sym_ref);
+  prt->idx_buffer_ref.val = JS_NULL;
+  prt->on_exception_ref.val = JS_NULL;
+  prt->message_handle_ref.val = JS_NULL;
+  prt->unneeded_ref.val = JS_NULL;
+  prt->actor_sym_ref.val = JS_NULL;
+
   cell_rt *crt = JS_GetContextOpaque(js);
   JS_FreeValue(js, js_blob_use(js));
 
-  // Load and compile engine.cm
-  size_t engine_size;
-  char *data = load_core_file(ENGINE, &engine_size);
-  if (!data) {
-    printf("ERROR: Could not load %s from %s!\n", ENGINE, core_path);
-    return;
+  // Load pre-compiled bootstrap bytecode (.mach), fall back to AST JSON
+  size_t boot_size;
+  int boot_is_bin = 1;
+  char *boot_data = load_core_file(BOOTSTRAP_MACH, &boot_size);
+  if (!boot_data) {
+    boot_is_bin = 0;
+    boot_data = load_core_file(BOOTSTRAP_AST, &boot_size);
   }
-
-  JSValue bytecode = JS_Compile(js, data, engine_size, ENGINE);
-  free(data);
-  if (JS_IsException(bytecode)) {
-    uncaught_exception(js, bytecode);
+  if (!boot_data) {
+    printf("ERROR: Could not load bootstrap from %s!\n", core_path);
     return;
   }
 
@@ -186,8 +171,8 @@ void script_startup(cell_rt *prt)
   JS_SetPropertyStr(js, hidden_env, "nota", js_nota_use(js));
   JS_SetPropertyStr(js, hidden_env, "wota", js_wota_use(js));
 
-  crt->actor_sym = JS_NewObject(js);
-  JS_SetPropertyStr(js, hidden_env, "actorsym", JS_DupValue(js, crt->actor_sym));
+  crt->actor_sym_ref.val = JS_NewObject(js);
+  JS_SetPropertyStr(js, hidden_env, "actorsym", JS_DupValue(js, crt->actor_sym_ref.val));
 
   // Always set init (even if null)
   if (crt->init_wota) {
@@ -205,9 +190,19 @@ void script_startup(cell_rt *prt)
   // Stone the environment
   hidden_env = JS_Stone(js, hidden_env);
 
-  // Integrate and run
+  // Run through MACH VM
   crt->state = ACTOR_RUNNING;
-  JSValue v = JS_Integrate(js, bytecode, hidden_env);
+  JSValue v;
+  if (boot_is_bin) {
+    v = JS_RunMachBin(js, (const uint8_t *)boot_data, boot_size, hidden_env);
+    free(boot_data);
+  } else {
+    cJSON *ast = cJSON_Parse(boot_data);
+    free(boot_data);
+    if (!ast) { printf("ERROR: Failed to parse bootstrap AST\n"); return; }
+    v = JS_RunMachTree(js, ast, hidden_env);
+    cJSON_Delete(ast);
+  }
   uncaught_exception(js, v);
   crt->state = ACTOR_IDLE;
   set_actor_state(crt);
@@ -255,108 +250,26 @@ static int run_test_suite(size_t heap_size)
   return result;
 }
 
-/* Run an immediate script string */
-static int run_eval(const char *script_or_file, int print_bytecode, int use_bootstrap_env)
+static void print_usage(const char *prog)
 {
-  if (!find_cell_shop()) return 1;
-
-  /* Check if argument is a file path */
-  struct stat st;
-  char *script = NULL;
-  char *allocated_script = NULL;
-  const char *filename = "<eval>";
-
-  if (stat(script_or_file, &st) == 0 && S_ISREG(st.st_mode)) {
-    /* It's a file, read its contents */
-    FILE *f = fopen(script_or_file, "r");
-    if (!f) {
-      printf("Failed to open file: %s\n", script_or_file);
-      return 1;
-    }
-    allocated_script = malloc(st.st_size + 1);
-    if (!allocated_script) {
-      fclose(f);
-      printf("Failed to allocate memory for script\n");
-      return 1;
-    }
-    size_t read_size = fread(allocated_script, 1, st.st_size, f);
-    fclose(f);
-    allocated_script[read_size] = '\0';
-    script = allocated_script;
-    filename = script_or_file;
-  } else {
-    /* Treat as inline script */
-    script = (char *)script_or_file;
-  }
-
-  JSRuntime *rt = JS_NewRuntime();
-  if (!rt) {
-    printf("Failed to create JS runtime\n");
-    free(allocated_script);
-    return 1;
-  }
-
-  JSContext *ctx = JS_NewContext(rt);
-  if (!ctx) {
-    printf("Failed to create JS context\n");
-    JS_FreeRuntime(rt);
-    free(allocated_script);
-    return 1;
-  }
-
-  int result = 0;
-
-  JSGCRef bytecode_ref;
-  JS_PushGCRef(ctx, &bytecode_ref);
-  bytecode_ref.val = JS_Compile(ctx, script, strlen(script), filename);
-  if (JS_IsException(bytecode_ref.val)) {
-    uncaught_exception(ctx, bytecode_ref.val);
-    JS_PopGCRef(ctx, &bytecode_ref);
-    result = 1;
-  } else {
-    if (print_bytecode) {
-      printf("=== Compiled Bytecode ===\n");
-      JS_DumpFunctionBytecode(ctx, bytecode_ref.val);
-    }
-    JSValue env = JS_NULL;
-    if (use_bootstrap_env) {
-      JSGCRef env_ref, json_ref, nota_ref, wota_ref;
-      JS_PushGCRef(ctx, &env_ref);
-      JS_PushGCRef(ctx, &json_ref);
-      JS_PushGCRef(ctx, &nota_ref);
-      JS_PushGCRef(ctx, &wota_ref);
-      env_ref.val = JS_NewObject(ctx);
-      /* Create modules with GC rooting, then stone them */
-      json_ref.val = js_json_use(ctx);
-      nota_ref.val = js_nota_use(ctx);
-      wota_ref.val = js_wota_use(ctx);
-      JS_SetPropertyStr(ctx, env_ref.val, "json", JS_Stone(ctx, json_ref.val));
-      JS_SetPropertyStr(ctx, env_ref.val, "nota", JS_Stone(ctx, nota_ref.val));
-      JS_SetPropertyStr(ctx, env_ref.val, "wota", JS_Stone(ctx, wota_ref.val));
-      env = JS_Stone(ctx, env_ref.val);
-      JS_PopGCRef(ctx, &wota_ref);
-      JS_PopGCRef(ctx, &nota_ref);
-      JS_PopGCRef(ctx, &json_ref);
-      JS_PopGCRef(ctx, &env_ref);
-    }
-    JSValue v = JS_Integrate(ctx, bytecode_ref.val, env);
-    JS_PopGCRef(ctx, &bytecode_ref);
-    if (JS_IsException(v)) {
-      uncaught_exception(ctx, v);
-      result = 1;
-    } else {
-      JS_FreeValue(ctx, v);
-    }
-  }
-
-  JS_FreeContext(ctx);
-  JS_FreeRuntime(rt);
-  free(allocated_script);
-  return result;
+  printf("Usage: %s [options] <script> [args...]\n\n", prog);
+  printf("Run a cell script (.ce actor or .cm module).\n\n");
+  printf("Options:\n");
+  printf("  --mcode <script> [args]    Run through mcode compilation pipeline\n");
+  printf("  --test [heap_size]         Run C test suite\n");
+  printf("  -h, --help                 Show this help message\n");
+  printf("\nRecompile after changes:  make\n");
+  printf("Bootstrap from scratch:  make bootstrap\n");
 }
 
 int cell_init(int argc, char **argv)
 {
+  /* Check for --help flag */
+  if (argc >= 2 && (strcmp(argv[1], "-h") == 0 || strcmp(argv[1], "--help") == 0)) {
+    print_usage(argv[0]);
+    return 0;
+  }
+
   /* Check for --test flag to run C test suite */
   if (argc >= 2 && strcmp(argv[1], "--test") == 0) {
     size_t heap_size = 64 * 1024; /* 64KB default */
@@ -370,451 +283,82 @@ int cell_init(int argc, char **argv)
     return run_test_suite(heap_size);
   }
 
-  /* Check for --ast flag to output AST JSON */
-  if (argc >= 3 && strcmp(argv[1], "--ast") == 0) {
-    const char *script_or_file = argv[2];
-    char *script = NULL;
-    char *allocated_script = NULL;
-    const char *filename = "<eval>";
-
-    struct stat st;
-    if (stat(script_or_file, &st) == 0 && S_ISREG(st.st_mode)) {
-      FILE *f = fopen(script_or_file, "r");
-      if (!f) {
-        printf("Failed to open file: %s\n", script_or_file);
-        return 1;
-      }
-      allocated_script = malloc(st.st_size + 1);
-      if (!allocated_script) {
-        fclose(f);
-        printf("Failed to allocate memory for script\n");
-        return 1;
-      }
-      size_t read_size = fread(allocated_script, 1, st.st_size, f);
-      fclose(f);
-      allocated_script[read_size] = '\0';
-      script = allocated_script;
-      filename = script_or_file;
-    } else {
-      script = (char *)script_or_file;
-    }
-
-    JSRuntime *rt = JS_NewRuntime();
-    if (!rt) {
-      printf("Failed to create JS runtime\n");
-      free(allocated_script);
-      return 1;
-    }
-    JSContext *ctx = JS_NewContext(rt);
-    if (!ctx) {
-      printf("Failed to create JS context\n");
-      JS_FreeRuntime(rt);
-      free(allocated_script);
-      return 1;
-    }
-
-    char *json = JS_AST(ctx, script, strlen(script), filename);
-    if (json) {
-      int has_errors = print_json_errors(json);
-      printf("%s\n", json);
-      free(json);
-      JS_FreeContext(ctx);
-      JS_FreeRuntime(rt);
-      free(allocated_script);
-      return has_errors ? 1 : 0;
-    } else {
-      printf("Failed to parse AST\n");
-      JS_FreeContext(ctx);
-      JS_FreeRuntime(rt);
-      free(allocated_script);
-      return 1;
-    }
+  /* Default: run script through bootstrap pipeline */
+  int use_mcode = 0;
+  int arg_start = 1;
+  if (argc >= 3 && strcmp(argv[1], "--mcode") == 0) {
+    use_mcode = 1;
+    arg_start = 2;
   }
 
-  /* Check for --tokenize flag to output token array JSON */
-  if (argc >= 3 && strcmp(argv[1], "--tokenize") == 0) {
-    const char *script_or_file = argv[2];
-    char *script = NULL;
-    char *allocated_script = NULL;
-    const char *filename = "<eval>";
+  if (!find_cell_shop()) return 1;
 
-    struct stat st;
-    if (stat(script_or_file, &st) == 0 && S_ISREG(st.st_mode)) {
-      FILE *f = fopen(script_or_file, "r");
-      if (!f) {
-        printf("Failed to open file: %s\n", script_or_file);
-        return 1;
-      }
-      allocated_script = malloc(st.st_size + 1);
-      if (!allocated_script) {
-        fclose(f);
-        printf("Failed to allocate memory for script\n");
-        return 1;
-      }
-      size_t read_size = fread(allocated_script, 1, st.st_size, f);
-      fclose(f);
-      allocated_script[read_size] = '\0';
-      script = allocated_script;
-      filename = script_or_file;
-    } else {
-      script = (char *)script_or_file;
-    }
-
-    JSRuntime *rt = JS_NewRuntime();
-    if (!rt) {
-      printf("Failed to create JS runtime\n");
-      free(allocated_script);
-      return 1;
-    }
-    JSContext *ctx = JS_NewContext(rt);
-    if (!ctx) {
-      printf("Failed to create JS context\n");
-      JS_FreeRuntime(rt);
-      free(allocated_script);
-      return 1;
-    }
-
-    char *json = JS_Tokenize(ctx, script, strlen(script), filename);
-    if (json) {
-      int has_errors = print_json_errors(json);
-      printf("%s\n", json);
-      free(json);
-      JS_FreeContext(ctx);
-      JS_FreeRuntime(rt);
-      free(allocated_script);
-      return has_errors ? 1 : 0;
-    } else {
-      printf("Failed to tokenize\n");
-      JS_FreeContext(ctx);
-      JS_FreeRuntime(rt);
-      free(allocated_script);
-      return 1;
-    }
+  size_t boot_size;
+  int boot_is_bin = 1;
+  char *boot_data = load_core_file(BOOTSTRAP_MACH, &boot_size);
+  if (!boot_data) {
+    boot_is_bin = 0;
+    boot_data = load_core_file(BOOTSTRAP_AST, &boot_size);
   }
-
-  /* Check for --mach flag to output machine code JSON */
-  if (argc >= 3 && strcmp(argv[1], "--mach") == 0) {
-    const char *script_or_file = argv[2];
-    char *script = NULL;
-    char *allocated_script = NULL;
-    const char *filename = "<eval>";
-
-    struct stat st;
-    if (stat(script_or_file, &st) == 0 && S_ISREG(st.st_mode)) {
-      FILE *f = fopen(script_or_file, "r");
-      if (!f) {
-        printf("Failed to open file: %s\n", script_or_file);
-        return 1;
-      }
-      allocated_script = malloc(st.st_size + 1);
-      if (!allocated_script) {
-        fclose(f);
-        printf("Failed to allocate memory for script\n");
-        return 1;
-      }
-      size_t read_size = fread(allocated_script, 1, st.st_size, f);
-      fclose(f);
-      allocated_script[read_size] = '\0';
-      script = allocated_script;
-      filename = script_or_file;
-    } else {
-      script = (char *)script_or_file;
-    }
-
-    JSRuntime *rt = JS_NewRuntime();
-    if (!rt) {
-      printf("Failed to create JS runtime\n");
-      free(allocated_script);
-      return 1;
-    }
-    JSContext *ctx = JS_NewContext(rt);
-    if (!ctx) {
-      printf("Failed to create JS context\n");
-      JS_FreeRuntime(rt);
-      free(allocated_script);
-      return 1;
-    }
-
-    char *ast_json = JS_AST(ctx, script, strlen(script), filename);
-    if (!ast_json) {
-      printf("Failed to parse AST\n");
-      JS_FreeContext(ctx);
-      JS_FreeRuntime(rt);
-      free(allocated_script);
-      return 1;
-    }
-
-    if (print_json_errors(ast_json)) {
-      free(ast_json);
-      JS_FreeContext(ctx);
-      JS_FreeRuntime(rt);
-      free(allocated_script);
-      return 1;
-    }
-
-    char *mach_json = JS_Mach(ctx, ast_json);
-    free(ast_json);
-
-    if (mach_json) {
-      if (print_json_errors(mach_json)) {
-        free(mach_json);
-        JS_FreeContext(ctx);
-        JS_FreeRuntime(rt);
-        free(allocated_script);
-        return 1;
-      }
-      printf("%s\n", mach_json);
-      free(mach_json);
-    } else {
-      printf("Failed to generate machine code\n");
-    }
-
-    JS_FreeContext(ctx);
-    JS_FreeRuntime(rt);
-    free(allocated_script);
-    return mach_json ? 0 : 1;
-  }
-
-  /* Check for --vmcode flag to dump linked register VM bytecode */
-  if (argc >= 3 && strcmp(argv[1], "--vmcode") == 0) {
-    const char *script_or_file = argv[2];
-    char *script = NULL;
-    char *allocated_script = NULL;
-    const char *filename = "<eval>";
-
-    struct stat st;
-    if (stat(script_or_file, &st) == 0 && S_ISREG(st.st_mode)) {
-      FILE *f = fopen(script_or_file, "r");
-      if (!f) {
-        printf("Failed to open file: %s\n", script_or_file);
-        return 1;
-      }
-      allocated_script = malloc(st.st_size + 1);
-      if (!allocated_script) {
-        fclose(f);
-        printf("Failed to allocate memory for script\n");
-        return 1;
-      }
-      size_t read_size = fread(allocated_script, 1, st.st_size, f);
-      fclose(f);
-      allocated_script[read_size] = '\0';
-      script = allocated_script;
-      filename = script_or_file;
-    } else {
-      script = (char *)script_or_file;
-    }
-
-    JSRuntime *rt = JS_NewRuntime();
-    if (!rt) {
-      printf("Failed to create JS runtime\n");
-      free(allocated_script);
-      return 1;
-    }
-    JSContext *ctx = JS_NewContext(rt);
-    if (!ctx) {
-      printf("Failed to create JS context\n");
-      JS_FreeRuntime(rt);
-      free(allocated_script);
-      return 1;
-    }
-
-    char *ast_json = JS_AST(ctx, script, strlen(script), filename);
-    if (!ast_json) {
-      printf("Failed to parse AST\n");
-      JS_FreeContext(ctx);
-      JS_FreeRuntime(rt);
-      free(allocated_script);
-      return 1;
-    }
-
-    if (print_json_errors(ast_json)) {
-      free(ast_json);
-      JS_FreeContext(ctx);
-      JS_FreeRuntime(rt);
-      free(allocated_script);
-      return 1;
-    }
-
-    char *mach_json = JS_Mach(ctx, ast_json);
-    free(ast_json);
-
-    if (mach_json) {
-      if (print_json_errors(mach_json)) {
-        free(mach_json);
-        JS_FreeContext(ctx);
-        JS_FreeRuntime(rt);
-        free(allocated_script);
-        return 1;
-      }
-      JS_DumpRegisterMach(ctx, mach_json, JS_NULL);
-      free(mach_json);
-    } else {
-      printf("Failed to generate machine code\n");
-    }
-
-    JS_FreeContext(ctx);
-    JS_FreeRuntime(rt);
-    free(allocated_script);
-    return mach_json ? 0 : 1;
-  }
-
-  /* Check for --mach-run flag to generate and run machine code through register VM */
-  if (argc >= 3 && strcmp(argv[1], "--mach-run") == 0) {
-    const char *script_or_file = argv[2];
-    char *script = NULL;
-    char *allocated_script = NULL;
-    const char *filename = "<eval>";
-
-    struct stat st;
-    if (stat(script_or_file, &st) == 0 && S_ISREG(st.st_mode)) {
-      FILE *f = fopen(script_or_file, "r");
-      if (!f) {
-        printf("Failed to open file: %s\n", script_or_file);
-        return 1;
-      }
-      allocated_script = malloc(st.st_size + 1);
-      if (!allocated_script) {
-        fclose(f);
-        printf("Failed to allocate memory for script\n");
-        return 1;
-      }
-      size_t read_size = fread(allocated_script, 1, st.st_size, f);
-      fclose(f);
-      allocated_script[read_size] = '\0';
-      script = allocated_script;
-      filename = script_or_file;
-    } else {
-      script = (char *)script_or_file;
-    }
-
-    JSRuntime *rt = JS_NewRuntime();
-    if (!rt) {
-      printf("Failed to create JS runtime\n");
-      free(allocated_script);
-      return 1;
-    }
-    JSContext *ctx = JS_NewContext(rt);
-    if (!ctx) {
-      printf("Failed to create JS context\n");
-      JS_FreeRuntime(rt);
-      free(allocated_script);
-      return 1;
-    }
-
-    /* Parse to AST */
-    char *ast_json = JS_AST(ctx, script, strlen(script), filename);
-    if (!ast_json) {
-      printf("Failed to parse AST\n");
-      JS_FreeContext(ctx);
-      JS_FreeRuntime(rt);
-      free(allocated_script);
-      return 1;
-    }
-
-    if (print_json_errors(ast_json)) {
-      free(ast_json);
-      JS_FreeContext(ctx);
-      JS_FreeRuntime(rt);
-      free(allocated_script);
-      return 1;
-    }
-
-    /* Generate machine code */
-    char *mach_json = JS_Mach(ctx, ast_json);
-    free(ast_json);
-
-    if (!mach_json) {
-      printf("Failed to generate machine code\n");
-      JS_FreeContext(ctx);
-      JS_FreeRuntime(rt);
-      free(allocated_script);
-      return 1;
-    }
-
-    if (print_json_errors(mach_json)) {
-      free(mach_json);
-      JS_FreeContext(ctx);
-      JS_FreeRuntime(rt);
-      free(allocated_script);
-      return 1;
-    }
-
-    /* Execute through register VM */
-    JSValue result = JS_IntegrateRegister(ctx, mach_json, JS_NULL);
-    free(mach_json);
-
-    int exit_code = 0;
-    if (JS_IsException(result)) {
-      JSValue exc = JS_GetException(ctx);
-      const char *err_str = JS_ToCString(ctx, exc);
-      if (err_str) {
-        printf("Error: %s\n", err_str);
-        JS_FreeCString(ctx, err_str);
-      }
-      exit_code = 1;
-    } else if (!JS_IsNull(result)) {
-      /* Print result */
-      const char *str = JS_ToCString(ctx, result);
-      if (str) {
-        printf("%s\n", str);
-        JS_FreeCString(ctx, str);
-      }
-    }
-
-    JS_FreeContext(ctx);
-    JS_FreeRuntime(rt);
-    free(allocated_script);
-    return exit_code;
-  }
-
-  /* Check for -e or --eval flag to run immediate script */
-  /* Also check for -p flag to print bytecode */
-  /* -s / --serializers flag provides json, nota, wota in env */
-  if (argc >= 3 && (strcmp(argv[1], "-e") == 0 || strcmp(argv[1], "--eval") == 0)) {
-    return run_eval(argv[2], 0, 0);
-  }
-  if (argc >= 3 && (strcmp(argv[1], "-p") == 0 || strcmp(argv[1], "--print-bytecode") == 0)) {
-    return run_eval(argv[2], 1, 0);
-  }
-  if (argc >= 3 && (strcmp(argv[1], "-s") == 0 || strcmp(argv[1], "--serializers") == 0)) {
-    return run_eval(argv[2], 0, 1);
-  }
-
-  int script_start = 1;
-
-  /* Find the cell shop at ~/.cell */
-  int found = find_cell_shop();
-  if (!found) {
+  if (!boot_data) {
+    printf("ERROR: Could not load bootstrap from %s\n", core_path);
     return 1;
   }
 
-  /* Create the initial actor from the command line */
-  int actor_argc = argc - script_start;
-  char **actor_argv = argv + script_start;
+  JSRuntime *rt = JS_NewRuntime();
+  if (!rt) {
+    printf("Failed to create JS runtime\n");
+    free(boot_data);
+    return 1;
+  }
+  JSContext *ctx = JS_NewContextWithHeapSize(rt, 16 * 1024 * 1024);
+  if (!ctx) {
+    printf("Failed to create JS context\n");
+    free(boot_data); JS_FreeRuntime(rt);
+    return 1;
+  }
 
-  WotaBuffer startwota;
-  wota_buffer_init(&startwota, 5);
-  wota_write_record(&startwota, 2);
-  wota_write_text(&startwota, "program");
-  wota_write_text(&startwota, actor_argv[0]);
-  wota_write_text(&startwota, "arg");
-  wota_write_array(&startwota, actor_argc - 1);
-  for (int i = 1; i < actor_argc; i++)
-    wota_write_text(&startwota, actor_argv[i]);
-    
-  /* Initialize synchronization primitives */
-  actor_initialize();
-    
-  root_cell = create_actor(startwota.data);
-#ifndef TARGET_PLAYDATE
-  signal(SIGINT, signal_handler);
-  signal(SIGTERM, signal_handler);
-  signal(SIGSEGV, signal_handler);
-  signal(SIGABRT, signal_handler);
-#endif  
-  actor_loop();
+  JS_FreeValue(ctx, js_blob_use(ctx));
 
-  return 0;
+  JSValue hidden_env = JS_NewObject(ctx);
+  JS_SetPropertyStr(ctx, hidden_env, "os", js_os_use(ctx));
+  JS_SetPropertyStr(ctx, hidden_env, "core_path", JS_NewString(ctx, core_path));
+  JS_SetPropertyStr(ctx, hidden_env, "use_mcode", JS_NewBool(ctx, use_mcode));
+  JSValue args_arr = JS_NewArray(ctx);
+  for (int i = arg_start; i < argc; i++) {
+    JSValue str = JS_NewString(ctx, argv[i]);
+    JS_ArrayPush(ctx, &args_arr, str);
+  }
+  JS_SetPropertyStr(ctx, hidden_env, "args", args_arr);
+  hidden_env = JS_Stone(ctx, hidden_env);
+
+  JSValue result;
+  if (boot_is_bin) {
+    result = JS_RunMachBin(ctx, (const uint8_t *)boot_data, boot_size, hidden_env);
+    free(boot_data);
+  } else {
+    cJSON *ast = cJSON_Parse(boot_data);
+    free(boot_data);
+    if (!ast) { printf("Failed to parse bootstrap AST\n"); JS_FreeContext(ctx); JS_FreeRuntime(rt); return 1; }
+    result = JS_RunMachTree(ctx, ast, hidden_env);
+    cJSON_Delete(ast);
+  }
+
+  int exit_code = 0;
+  if (JS_IsException(result)) {
+    JS_GetException(ctx);
+    exit_code = 1;
+  } else if (!JS_IsNull(result)) {
+    const char *str = JS_ToCString(ctx, result);
+    if (str) {
+      printf("%s\n", str);
+      JS_FreeCString(ctx, str);
+    }
+  }
+
+  JS_FreeContext(ctx);
+  JS_FreeRuntime(rt);
+  return exit_code;
 }
 
 int JS_ArrayLength(JSContext *js, JSValue a)
@@ -848,36 +392,16 @@ double cell_random() {
 
 void cell_trace_sethook(cell_hook)
 {
-  
+
 }
 
 int uncaught_exception(JSContext *js, JSValue v)
 {
-  cell_rt *rt = JS_GetContextOpaque(js);
-  if (!JS_HasException(js)) {
-    JS_FreeValue(js,v);
+  (void)v;
+  if (!JS_HasException(js))
     return 1;
-  }
-
-  JSValue exp = JS_GetException(js);
-  
-  JSValue message = JS_GetPropertyStr(js, exp, "message");
-  const char *msg_str = JS_ToCString(js, message);
-  if (msg_str) {
-    printf("Exception: %s\n", msg_str);
-    JS_FreeCString(js, msg_str);
-  }
-  JS_FreeValue(js, message);
-  
-  JSValue stack = JS_GetPropertyStr(js, exp, "stack");
-  const char *stack_str = JS_ToCString(js, stack);
-  if (stack_str) {
-    printf("Stack:\n%s\n", stack_str);
-    JS_FreeCString(js, stack_str);
-  }
-  JS_FreeValue(js, stack);
-  
-  JS_FreeValue(js, exp);
-  JS_FreeValue(js, v);
+  /* Error message and backtrace were already printed to stderr
+     by JS_ThrowError2 / print_backtrace. Just clear the flag. */
+  JS_GetException(js);
   return 0;
-}
+}

source/cell.h

@@ -144,7 +144,7 @@ JS_SetPropertyFunctionList(js, TYPE##_proto, js_##TYPE##_funcs, countof(js_##TYP
 
 #define QJSCLASSPREP_NO_FUNCS(TYPE) \
 JS_NewClassID(&js_##TYPE##_id);\
-JS_NewClass(JS_GetRuntime(js), js_##TYPE##_id, &js_##TYPE##_class);\
+JS_NewClass(js, js_##TYPE##_id, &js_##TYPE##_class);\
 JSValue TYPE##_proto = JS_NewObject(js); \
 JS_SetClassProto(js, js_##TYPE##_id, TYPE##_proto); \
 

source/cell_internal.h

@@ -24,21 +24,26 @@ typedef struct letter {
 
 typedef struct cell_rt {
   JSContext *context;
-  JSValue idx_buffer;
-  JSValue on_exception;
-  JSValue message_handle;
+
+  /* JSValues on the GC heap — each paired with a JSGCRef so the
+     Cheney GC can relocate them during compaction. */
+  JSGCRef idx_buffer_ref;
+  JSGCRef on_exception_ref;
+  JSGCRef message_handle_ref;
+  JSGCRef unneeded_ref;
+  JSGCRef actor_sym_ref;
 
   void *init_wota;
 
   /* Protects JSContext usage */
   pthread_mutex_t *mutex; /* for everything else */
   pthread_mutex_t *msg_mutex; /* For message queue and timers queue */
-  
+
   char *id;
 
   int idx_count;
 
-  /* The “mailbox” for incoming messages + a dedicated lock for it: */
+  /* The "mailbox" for incoming messages + a dedicated lock for it: */
   letter *letters;
 
   /* CHANGED FOR EVENTS: a separate lock for the actor->events queue */
@@ -47,14 +52,11 @@ typedef struct cell_rt {
   int state;
   uint32_t ar;  // timer for unneeded
   double ar_secs;  // time for unneeded
-  JSValue unneeded;  // fn to call before unneeded
-    
+
   int disrupt;
   int main_thread_only;
   int affinity;
-  
-  JSValue actor_sym;
-  
+
   const char *name; // human friendly name
   cell_hook trace_hook;
 } cell_rt;
@@ -63,8 +65,6 @@ cell_rt *create_actor(void *wota);
 const char *register_actor(const char *id, cell_rt *actor, int mainthread, double ar);
 void actor_disrupt(cell_rt *actor);
 
-JSValue actor_sym(cell_rt *actor);
-
 const char *send_message(const char *id, void *msg);
 const char *register_actor(const char *id, cell_rt *actor, int mainthread, double ar);
 void actor_unneeded(cell_rt *actor, JSValue fn, double seconds);

source/qbe_helpers.c

@@ -0,0 +1,194 @@
+/*
+ * QBE Helper Functions
+ *
+ * Thin C wrappers called from QBE-generated code for operations
+ * that are too complex to inline: float arithmetic, float comparison,
+ * string comparison, bitwise ops on floats, and boolean conversion.
+ */
+
+#include "quickjs-internal.h"
+#include <math.h>
+
+/* Comparison op IDs (must match qbe.cm float_cmp_op_id values) */
+enum {
+  QBE_CMP_EQ = 0,
+  QBE_CMP_NE = 1,
+  QBE_CMP_LT = 2,
+  QBE_CMP_LE = 3,
+  QBE_CMP_GT = 4,
+  QBE_CMP_GE = 5
+};
+
+/* ============================================================
+   Float binary arithmetic
+   ============================================================ */
+
+static inline JSValue qbe_float_binop(JSContext *ctx, JSValue a, JSValue b,
+                                      double (*op)(double, double)) {
+  double da, db;
+  JS_ToFloat64(ctx, &da, a);
+  JS_ToFloat64(ctx, &db, b);
+  double r = op(da, db);
+  if (!isfinite(r))
+    return JS_NULL;
+  return JS_NewFloat64(ctx, r);
+}
+
+static double op_add(double a, double b) { return a + b; }
+static double op_sub(double a, double b) { return a - b; }
+static double op_mul(double a, double b) { return a * b; }
+
+JSValue qbe_float_add(JSContext *ctx, JSValue a, JSValue b) {
+  return qbe_float_binop(ctx, a, b, op_add);
+}
+
+JSValue qbe_float_sub(JSContext *ctx, JSValue a, JSValue b) {
+  return qbe_float_binop(ctx, a, b, op_sub);
+}
+
+JSValue qbe_float_mul(JSContext *ctx, JSValue a, JSValue b) {
+  return qbe_float_binop(ctx, a, b, op_mul);
+}
+
+JSValue qbe_float_div(JSContext *ctx, JSValue a, JSValue b) {
+  double da, db;
+  JS_ToFloat64(ctx, &da, a);
+  JS_ToFloat64(ctx, &db, b);
+  if (db == 0.0)
+    return JS_NULL;
+  double r = da / db;
+  if (!isfinite(r))
+    return JS_NULL;
+  return JS_NewFloat64(ctx, r);
+}
+
+JSValue qbe_float_mod(JSContext *ctx, JSValue a, JSValue b) {
+  double da, db;
+  JS_ToFloat64(ctx, &da, a);
+  JS_ToFloat64(ctx, &db, b);
+  if (db == 0.0)
+    return JS_NULL;
+  double r = fmod(da, db);
+  if (!isfinite(r))
+    return JS_NULL;
+  return JS_NewFloat64(ctx, r);
+}
+
+JSValue qbe_float_pow(JSContext *ctx, JSValue a, JSValue b) {
+  double da, db;
+  JS_ToFloat64(ctx, &da, a);
+  JS_ToFloat64(ctx, &db, b);
+  double r = pow(da, db);
+  if (!isfinite(r) && isfinite(da) && isfinite(db))
+    return JS_NULL;
+  return JS_NewFloat64(ctx, r);
+}
+
+/* ============================================================
+   Float unary ops
+   ============================================================ */
+
+JSValue qbe_float_neg(JSContext *ctx, JSValue v) {
+  double d;
+  JS_ToFloat64(ctx, &d, v);
+  return JS_NewFloat64(ctx, -d);
+}
+
+JSValue qbe_float_inc(JSContext *ctx, JSValue v) {
+  double d;
+  JS_ToFloat64(ctx, &d, v);
+  return JS_NewFloat64(ctx, d + 1);
+}
+
+JSValue qbe_float_dec(JSContext *ctx, JSValue v) {
+  double d;
+  JS_ToFloat64(ctx, &d, v);
+  return JS_NewFloat64(ctx, d - 1);
+}
+
+/* ============================================================
+   Float comparison — returns 0 or 1 for QBE branching
+   ============================================================ */
+
+int qbe_float_cmp(JSContext *ctx, int op, JSValue a, JSValue b) {
+  double da, db;
+  JS_ToFloat64(ctx, &da, a);
+  JS_ToFloat64(ctx, &db, b);
+  switch (op) {
+  case QBE_CMP_EQ: return da == db;
+  case QBE_CMP_NE: return da != db;
+  case QBE_CMP_LT: return da < db;
+  case QBE_CMP_LE: return da <= db;
+  case QBE_CMP_GT: return da > db;
+  case QBE_CMP_GE: return da >= db;
+  default: return 0;
+  }
+}
+
+/* ============================================================
+   Boolean conversion wrapper
+   ============================================================ */
+
+int qbe_to_bool(JSContext *ctx, JSValue v) {
+  return JS_ToBool(ctx, v);
+}
+
+/* ============================================================
+   Bitwise not on non-int (float -> int32 -> ~)
+   ============================================================ */
+
+JSValue qbe_bnot(JSContext *ctx, JSValue v) {
+  int32_t i;
+  JS_ToInt32(ctx, &i, v);
+  return JS_NewInt32(ctx, ~i);
+}
+
+/* ============================================================
+   Bitwise binary ops on floats (convert both to int32, apply, re-tag)
+   ============================================================ */
+
+JSValue qbe_bitwise_and(JSContext *ctx, JSValue a, JSValue b) {
+  int32_t ia, ib;
+  JS_ToInt32(ctx, &ia, a);
+  JS_ToInt32(ctx, &ib, b);
+  return JS_NewInt32(ctx, ia & ib);
+}
+
+JSValue qbe_bitwise_or(JSContext *ctx, JSValue a, JSValue b) {
+  int32_t ia, ib;
+  JS_ToInt32(ctx, &ia, a);
+  JS_ToInt32(ctx, &ib, b);
+  return JS_NewInt32(ctx, ia | ib);
+}
+
+JSValue qbe_bitwise_xor(JSContext *ctx, JSValue a, JSValue b) {
+  int32_t ia, ib;
+  JS_ToInt32(ctx, &ia, a);
+  JS_ToInt32(ctx, &ib, b);
+  return JS_NewInt32(ctx, ia ^ ib);
+}
+
+/* ============================================================
+   Shift ops on floats (convert to int32, shift, re-tag)
+   ============================================================ */
+
+JSValue qbe_shift_shl(JSContext *ctx, JSValue a, JSValue b) {
+  int32_t ia, ib;
+  JS_ToInt32(ctx, &ia, a);
+  JS_ToInt32(ctx, &ib, b);
+  return JS_NewInt32(ctx, ia << (ib & 31));
+}
+
+JSValue qbe_shift_sar(JSContext *ctx, JSValue a, JSValue b) {
+  int32_t ia, ib;
+  JS_ToInt32(ctx, &ia, a);
+  JS_ToInt32(ctx, &ib, b);
+  return JS_NewInt32(ctx, ia >> (ib & 31));
+}
+
+JSValue qbe_shift_shr(JSContext *ctx, JSValue a, JSValue b) {
+  int32_t ia, ib;
+  JS_ToInt32(ctx, &ia, a);
+  JS_ToInt32(ctx, &ib, b);
+  return JS_NewInt32(ctx, (uint32_t)ia >> (ib & 31));
+}

source/qjs_actor.c

@@ -74,7 +74,7 @@ JSC_CCALL(os_register_actor,
   JS_ToFloat64(js, &ar, argv[3]);
   const char *err = register_actor(id, rt, JS_ToBool(js, argv[2]), ar);
   if (err) return JS_ThrowInternalError(js, "Could not register actor: %s", err);
-  rt->message_handle = JS_DupValue(js, argv[1]);
+  rt->message_handle_ref.val = argv[1];
   rt->context = js;
   JS_FreeCString(js, id);
 )
@@ -106,8 +106,7 @@ JSC_SCALL(actor_setname,
 
 JSC_CCALL(actor_on_exception,
   cell_rt *rt = JS_GetContextOpaque(js);
-  JS_FreeValue(js, rt->on_exception);
-  rt->on_exception = JS_DupValue(js,argv[0]);
+  rt->on_exception_ref.val = argv[0];
 )
 
 JSC_CCALL(actor_clock,

source/quickjs.h

@@ -367,10 +367,10 @@ JSRuntime *JS_NewRuntime (void);
 void JS_SetRuntimeInfo (JSRuntime *rt, const char *info);
 void JS_SetMemoryLimit (JSRuntime *rt, size_t limit);
 /* use 0 to disable maximum stack size check */
-void JS_SetMaxStackSize (JSRuntime *rt, size_t stack_size);
+void JS_SetMaxStackSize (JSContext *ctx, size_t stack_size);
 /* should be called when changing thread to update the stack top value
    used to check stack overflow. */
-void JS_UpdateStackTop (JSRuntime *rt);
+void JS_UpdateStackTop (JSContext *ctx);
 void JS_FreeRuntime (JSRuntime *rt);
 void *JS_GetRuntimeOpaque (JSRuntime *rt);
 void JS_SetRuntimeOpaque (JSRuntime *rt, void *opaque);
@@ -429,9 +429,9 @@ JSClassID JS_NewClassID (JSClassID *pclass_id);
 /* Returns the class ID if `v` is an object, otherwise returns
  * JS_INVALID_CLASS_ID. */
 JSClassID JS_GetClassID (JSValue v);
-int JS_NewClass (JSRuntime *rt, JSClassID class_id,
+int JS_NewClass (JSContext *ctx, JSClassID class_id,
                  const JSClassDef *class_def);
-int JS_IsRegisteredClass (JSRuntime *rt, JSClassID class_id);
+int JS_IsRegisteredClass (JSContext *ctx, JSClassID class_id);
 
 extern JSClassID js_class_id_alloc;
 
@@ -577,11 +577,8 @@ static JS_BOOL JS_IsStone(JSValue v);
 int JS_GetLength (JSContext *ctx, JSValue obj, int64_t *pres);
 
 JSValue JS_Throw (JSContext *ctx, JSValue obj);
-void JS_SetUncatchableException (JSContext *ctx, JS_BOOL flag);
 JSValue JS_GetException (JSContext *ctx);
 JS_BOOL JS_HasException (JSContext *ctx);
-JS_BOOL JS_IsError (JSContext *ctx, JSValue val);
-JSValue JS_NewError (JSContext *ctx);
 JSValue __js_printf_like (2, 3)
   JS_ThrowSyntaxError (JSContext *ctx, const char *fmt, ...);
 JSValue __js_printf_like (2, 3)
@@ -697,6 +694,9 @@ JSValue JS_GetProperty (JSContext *ctx, JSValue this_obj, JSValue prop);
 // For records
 JSValue JS_GetPropertyStr (JSContext *ctx, JSValue this_obj, const char *prop);
 int JS_SetPropertyStr (JSContext *ctx, JSValue this_obj, const char *prop, JSValue val);
+
+// Set property on the global object
+int JS_SetGlobalStr (JSContext *ctx, const char *prop, JSValue val);
 int JS_SetProperty (JSContext *ctx, JSValue this_obj, JSValue prop, JSValue val);
 JSValue JS_GetPrototype (JSContext *ctx, JSValue val);
 
@@ -714,16 +714,6 @@ JSValue JS_GetOwnPropertyNames (JSContext *ctx, JSValue obj);
 
 JSValue JS_Call (JSContext *ctx, JSValue func_obj, JSValue this_obj, int argc, JSValue *argv);
 
-/* Compile source code to bytecode without executing.
-   'input' must be zero terminated i.e. input[input_len] = '\0'.
-   Returns unlinked bytecode on success, JS_EXCEPTION on error. */
-JSValue JS_Compile (JSContext *ctx, const char *input, size_t input_len,
-                    const char *filename);
-
-/* Link compiled bytecode with environment and execute.
-   env should be stoned record or null.
-   Variables resolve: env first, then global intrinsics. */
-JSValue JS_Integrate (JSContext *ctx, JSValue bytecode, JSValue env);
 void JS_SetOpaque (JSValue obj, void *opaque);
 void *JS_GetOpaque (JSValue obj, JSClassID class_id);
 void *JS_GetOpaque2 (JSContext *ctx, JSValue obj, JSClassID class_id);
@@ -740,7 +730,7 @@ JSValue JS_JSONStringify (JSContext *ctx, JSValue obj,
 
 /* return != 0 if the JS code needs to be interrupted */
 typedef int JSInterruptHandler (JSRuntime *rt, void *opaque);
-void JS_SetInterruptHandler (JSRuntime *rt, JSInterruptHandler *cb,
+void JS_SetInterruptHandler (JSContext *ctx, JSInterruptHandler *cb,
                              void *opaque);
 /* select which debug info is stripped from the compiled code */
 #define JS_STRIP_SOURCE (1 << 0) /* strip source code */
@@ -749,28 +739,6 @@ void JS_SetInterruptHandler (JSRuntime *rt, JSInterruptHandler *cb,
 void JS_SetStripInfo (JSRuntime *rt, int flags);
 int JS_GetStripInfo (JSRuntime *rt);
 
-/* Object Writer/Reader (currently only used to handle precompiled code) */
-#define JS_WRITE_OBJ_BYTECODE (1 << 0) /* allow function/module */
-#define JS_WRITE_OBJ_BSWAP (1 << 1)    /* byte swapped output */
-#define JS_WRITE_OBJ_SAB (1 << 2)      /* allow SharedArrayBuffer */
-#define JS_WRITE_OBJ_REFERENCE                                                \
-  (1 << 3) /* allow object references to                                      \
-              encode arbitrary object                                         \
-              graph */
-uint8_t *JS_WriteObject (JSContext *ctx, size_t *psize, JSValue obj,
-                         int flags);
-uint8_t *JS_WriteObject2 (JSContext *ctx, size_t *psize, JSValue obj,
-                          int flags, uint8_t ***psab_tab,
-                          size_t *psab_tab_len);
-
-#define JS_READ_OBJ_BYTECODE (1 << 0)  /* allow function/module */
-#define JS_READ_OBJ_ROM_DATA (1 << 1)  /* avoid duplicating 'buf' data */
-#define JS_READ_OBJ_SAB (1 << 2)       /* allow SharedArrayBuffer */
-#define JS_READ_OBJ_REFERENCE (1 << 3) /* allow object references */
-JSValue JS_ReadObject (JSContext *ctx, const uint8_t *buf, size_t buf_len,
-                       int flags);
-/* Dump bytecode of a compiled function (for debugging) */
-void JS_DumpFunctionBytecode (JSContext *ctx, JSValue func_val);
 
 /* C function definition */
 typedef enum JSCFunctionEnum {
@@ -1100,149 +1068,58 @@ void *js_malloc_rt (size_t size);
 void *js_mallocz_rt (size_t size);
 void js_free_rt (void *ptr);
 
-/* ============================================================================
-   Context-Neutral Module Format (CellModule)
-   ============================================================================ */
+struct cJSON;
 
-/* Capture descriptor - what a nested function closes over */
-typedef enum {
-  CAP_FROM_PARENT_LOCAL = 1,   /* capture local from parent function */
-  CAP_FROM_PARENT_UPVALUE = 2  /* forward upvalue from parent's upvalues */
-} CellCapKind;
+/* Compiled bytecode (context-free, serializable) */
+typedef struct MachCode MachCode;
 
-typedef struct CellCapDesc {
-  uint8_t kind;    /* CAP_FROM_PARENT_LOCAL or CAP_FROM_PARENT_UPVALUE */
-  uint16_t index;  /* local index in parent, or upvalue index in parent */
-} CellCapDesc;
+/* Compile AST cJSON tree to context-free MachCode. */
+MachCode *JS_CompileMachTree(struct cJSON *ast);
 
-/* External relocation - for integrate-time patching */
-typedef enum {
-  EXT_GET = 1,  /* OP_get_var -> OP_get_env_slot or OP_get_global_slot */
-  EXT_SET = 2   /* OP_put_var -> OP_set_env_slot or OP_set_global_slot */
-} CellExtKind;
+/* Compile AST JSON string to context-free MachCode. */
+MachCode *JS_CompileMach(const char *ast_json);
 
-typedef struct CellExternalReloc {
-  uint32_t pc_offset;  /* where operand lives in bytecode */
-  uint32_t name_sid;   /* string id of the external name */
-  uint8_t kind;        /* EXT_GET or EXT_SET */
-} CellExternalReloc;
+/* Free a compiled MachCode tree. */
+void JS_FreeMachCode(MachCode *mc);
 
-/* Constant types in cpool */
-typedef enum {
-  CELL_CONST_NULL = 0,
-  CELL_CONST_INT = 1,
-  CELL_CONST_FLOAT = 2,
-  CELL_CONST_STRING = 3,  /* string_sid into module string table */
-  CELL_CONST_UNIT = 4     /* unit_id for nested function */
-} CellConstType;
+/* Serialize MachCode to binary. Returns sys_malloc'd buffer; caller frees. */
+uint8_t *JS_SerializeMachCode(MachCode *mc, size_t *out_size);
 
-typedef struct CellConst {
-  uint8_t type;  /* CellConstType */
-  union {
-    int32_t i32;
-    double f64;
-    uint32_t string_sid;
-    uint32_t unit_id;
-  };
-} CellConst;
+/* Deserialize binary back to MachCode. Returns NULL on error. */
+MachCode *JS_DeserializeMachCode(const uint8_t *data, size_t size);
 
-/* Per-unit structure (context-neutral, flattened) */
-typedef struct CellUnit {
-  /* Constant pool */
-  uint32_t const_count;
-  CellConst *constants;
+/* Load compiled MachCode into a JSContext, materializing JSValues. */
+struct JSCodeRegister *JS_LoadMachCode(JSContext *ctx, MachCode *mc, JSValue env);
 
-  /* Bytecode */
-  uint32_t bytecode_len;
-  uint8_t *bytecode;
+/* Dump MACH bytecode to stdout. Takes AST cJSON tree. */
+void JS_DumpMachTree (JSContext *ctx, struct cJSON *ast, JSValue env);
 
-  /* Stack requirements */
-  uint16_t arg_count;
-  uint16_t var_count;
-  uint16_t stack_size;
+/* Dump MACH bytecode to stdout. Takes AST JSON string. */
+void JS_DumpMach (JSContext *ctx, const char *ast_json, JSValue env);
 
-  /* Upvalue (capture) descriptors */
-  uint16_t upvalue_count;
-  CellCapDesc *upvalues;
+/* Compile and execute MACH bytecode from AST cJSON tree. */
+JSValue JS_RunMachTree (JSContext *ctx, struct cJSON *ast, JSValue env);
 
-  /* External relocations */
-  uint32_t external_count;
-  CellExternalReloc *externals;
+/* Compile and execute MACH bytecode from AST JSON string. */
+JSValue JS_RunMach (JSContext *ctx, const char *ast_json, JSValue env);
 
-  /* Debug info (optional) */
-  uint32_t pc2line_len;
-  uint8_t *pc2line;
-  uint32_t name_sid;  /* unit name for stack traces */
-} CellUnit;
+/* Deserialize and execute pre-compiled MACH binary bytecode. */
+JSValue JS_RunMachBin(JSContext *ctx, const uint8_t *data, size_t size, JSValue env);
 
-/* Module-level structure (context-neutral) */
-#define CELL_MODULE_MAGIC 0x4C4C4543  /* "CELL" */
-#define CELL_MODULE_VERSION 1
+/* Execute MCODE from cJSON tree. Takes ownership of root. */
+JSValue JS_CallMcodeTree (JSContext *ctx, struct cJSON *root);
 
-typedef struct CellModule {
-  uint32_t magic;    /* CELL_MODULE_MAGIC */
-  uint8_t version;   /* CELL_MODULE_VERSION */
-  uint8_t flags;
+/* Execute MCODE from cJSON tree with hidden env. Takes ownership of root. */
+JSValue JS_CallMcodeTreeEnv (JSContext *ctx, struct cJSON *root, JSValue env);
 
-  /* Shared string table (module-global) */
-  uint32_t string_count;
-  uint32_t string_data_size;
-  uint8_t *string_data;      /* concatenated UTF-8 strings */
-  uint32_t *string_offsets;  /* offset for each string */
-
-  /* Unit table (entry 0 is the main/entry unit) */
-  uint32_t unit_count;
-  CellUnit *units;
-
-  /* Debug: source stored once at module level */
-  uint32_t source_len;
-  char *source;
-} CellModule;
-
-/* Free a CellModule and all its contents */
-void cell_module_free (CellModule *mod);
-
-/* Write a CellModule to a byte buffer.
-   Returns allocated buffer (caller must free with pjs_free), or NULL on error. */
-uint8_t *cell_module_write (CellModule *mod, size_t *out_len);
-
-/* Read a CellModule from a byte buffer.
-   Returns allocated CellModule (caller must free with cell_module_free), or NULL on error. */
-CellModule *cell_module_read (const uint8_t *buf, size_t buf_len);
-
-/* Convert compiled JSFunctionBytecode to CellModule.
-   Returns allocated CellModule (caller must free with cell_module_free), or NULL on error. */
-CellModule *cell_module_from_bytecode (JSContext *ctx, JSFunctionBytecode *main_func);
-
-/* Compile source code directly to CellModule.
-   Returns allocated CellModule (caller must free with cell_module_free), or NULL on error. */
-CellModule *JS_CompileModule (JSContext *ctx, const char *input, size_t input_len, const char *filename);
-
-/* Parse source code and return AST as JSON string.
-   Returns malloc'd JSON string (caller must free), or NULL on error. */
-char *JS_AST (JSContext *ctx, const char *source, size_t len, const char *filename);
-
-/* Tokenize source code and return token array as JSON string.
-   Returns malloc'd JSON string (caller must free), or NULL on error. */
-char *JS_Tokenize (JSContext *ctx, const char *source, size_t len, const char *filename);
-
-/* Generate register-based machine code from AST JSON.
-   Returns malloc'd JSON string (caller must free), or NULL on error. */
-char *JS_Mach (JSContext *ctx, const char *ast_json);
-
-/* Link and execute register-based machine code.
-   mach_json: JSON output from JS_Mach
-   env: stone record for external variable resolution (or JS_NULL)
+/* Parse and execute MCODE JSON string.
    Returns result of execution, or JS_EXCEPTION on error. */
-JSValue JS_IntegrateRegister (JSContext *ctx, const char *mach_json, JSValue env);
+JSValue JS_CallMcode (JSContext *ctx, const char *mcode_json);
 
-/* Dump linked register VM bytecode to stdout for debugging.
-   Links mach_json and prints the resulting bytecode with resolved labels. */
-void JS_DumpRegisterMach (JSContext *ctx, const char *mach_json, JSValue env);
-
-/* Integrate a CellModule with an environment and execute.
-   Returns callable function value, or JS_EXCEPTION on error. */
-JSValue cell_module_integrate (JSContext *ctx, CellModule *mod, JSValue env);
+/* Get stack trace as cJSON array of frame objects.
+   Returns NULL if no register VM frame is active.
+   Caller must call cJSON_Delete() on the result. */
+struct cJSON *JS_GetStack (JSContext *ctx);
 
 #undef js_unlikely
 #undef inline

source/scheduler.c

@@ -266,11 +266,11 @@ void actor_free(cell_rt *actor)
 
   JSContext *js = actor->context;
   
-  JS_FreeValue(js, actor->idx_buffer);
-  JS_FreeValue(js, actor->message_handle);
-  JS_FreeValue(js, actor->on_exception);
-  JS_FreeValue(js, actor->unneeded);
-  JS_FreeValue(js, actor->actor_sym);
+  JS_DeleteGCRef(js, &actor->idx_buffer_ref);
+  JS_DeleteGCRef(js, &actor->on_exception_ref);
+  JS_DeleteGCRef(js, &actor->message_handle_ref);
+  JS_DeleteGCRef(js, &actor->unneeded_ref);
+  JS_DeleteGCRef(js, &actor->actor_sym_ref);
   
   for (int i = 0; i < hmlen(actor->timers); i++) {
     JS_FreeValue(js, actor->timers[i].value);
@@ -289,10 +289,8 @@ void actor_free(cell_rt *actor)
   
   arrfree(actor->letters);
   
-  JSRuntime *rt = JS_GetRuntime(js);
-  JS_SetInterruptHandler(rt, NULL, NULL);  
+  JS_SetInterruptHandler(js, NULL, NULL);
   JS_FreeContext(js);
-  JS_FreeRuntime(rt);
   free(actor->id);
   
   pthread_mutex_unlock(actor->mutex);
@@ -419,8 +417,8 @@ uint32_t actor_remove_cb(cell_rt *actor, uint32_t id, uint32_t interval)
   
   actor->disrupt = 1;
   
-  if (!JS_IsNull(actor->unneeded)) {
-    JSValue ret = JS_Call(actor->context, actor->unneeded, JS_NULL, 0, NULL);
+  if (!JS_IsNull(actor->unneeded_ref.val)) {
+    JSValue ret = JS_Call(actor->context, actor->unneeded_ref.val, JS_NULL, 0, NULL);
     uncaught_exception(actor->context, ret);
   }
 
@@ -434,14 +432,13 @@ uint32_t actor_remove_cb(cell_rt *actor, uint32_t id, uint32_t interval)
 void actor_unneeded(cell_rt *actor, JSValue fn, double seconds)
 {
   if (actor->disrupt) return;
-  JS_FreeValue(actor->context, actor->unneeded);
-  
+
   if (!JS_IsFunction(fn)) {
-    actor->unneeded = JS_NULL;
+    actor->unneeded_ref.val = JS_NULL;
     goto END;
   }
-  
-  actor->unneeded = JS_DupValue(actor->context, fn);
+
+  actor->unneeded_ref.val = fn;
   actor->ar_secs = seconds;
   
   END:
@@ -493,11 +490,10 @@ cell_rt *create_actor(void *wota)
   actor->heap = mi_heap_new();
 #endif
   actor->init_wota = wota;
-  actor->idx_buffer = JS_NULL;
-  actor->message_handle = JS_NULL;
-  actor->unneeded = JS_NULL;
-  actor->on_exception = JS_NULL;
-  actor->actor_sym = JS_NULL;
+  /* GCRef fields are registered after JSContext creation in script_startup.
+     For now, zero-init from calloc is sufficient (val = 0 = JS_MKVAL(JS_TAG_INT,0),
+     which is not a pointer so GC-safe). The actual JS_NULL assignment and
+     JS_AddGCRef happen in script_startup. */
 
   arrsetcap(actor->letters, 5);
 
@@ -587,7 +583,7 @@ void actor_turn(cell_rt *actor)
     size_t size = blob_length(l.blob_data) / 8; // Convert bits to bytes
     JSValue arg = js_new_blob_stoned_copy(actor->context, (void*)blob_data(l.blob_data), size);
     blob_destroy(l.blob_data);
-    result = JS_Call(actor->context, actor->message_handle, JS_NULL, 1, &arg);
+    result = JS_Call(actor->context, actor->message_handle_ref.val, JS_NULL, 1, &arg);
     uncaught_exception(actor->context, result);
     JS_FreeValue(actor->context, arg);
   } else if (l.type == LETTER_CALLBACK) {

source/scheduler_playdate.c

@@ -118,11 +118,11 @@ void actor_free(cell_rt *actor)
   
   JSContext *js = actor->context;
   
-  JS_FreeValue(js, actor->idx_buffer);
-  JS_FreeValue(js, actor->message_handle);
-  JS_FreeValue(js, actor->on_exception);
-  JS_FreeValue(js, actor->unneeded);
-  JS_FreeValue(js, actor->actor_sym);
+  JS_DeleteGCRef(js, &actor->idx_buffer_ref);
+  JS_DeleteGCRef(js, &actor->on_exception_ref);
+  JS_DeleteGCRef(js, &actor->message_handle_ref);
+  JS_DeleteGCRef(js, &actor->unneeded_ref);
+  JS_DeleteGCRef(js, &actor->actor_sym_ref);
   
   /* Free timer callbacks stored in actor */
   for (int i = 0; i < hmlen(actor->timers); i++) {
@@ -142,10 +142,8 @@ void actor_free(cell_rt *actor)
   
   arrfree(actor->letters);
   
-  JSRuntime *rt = JS_GetRuntime(js);
-  JS_SetInterruptHandler(rt, NULL, NULL);  
+  JS_SetInterruptHandler(js, NULL, NULL);
   JS_FreeContext(js);
-  JS_FreeRuntime(rt);
   free(actor->id);
   
   free(actor);
@@ -157,14 +155,13 @@ void actor_free(cell_rt *actor)
 void actor_unneeded(cell_rt *actor, JSValue fn, double seconds)
 {
   if (actor->disrupt) return;
-  JS_FreeValue(actor->context, actor->unneeded);
-  
-  if (!JS_IsFunction(actor->context, fn)) {
-    actor->unneeded = JS_NULL;
+
+  if (!JS_IsFunction(fn)) {
+    actor->unneeded_ref.val = JS_NULL;
     goto END;
   }
-  
-  actor->unneeded = JS_DupValue(actor->context, fn);
+
+  actor->unneeded_ref.val = fn;
   actor->ar_secs = seconds;
   
   END:
@@ -257,8 +254,8 @@ uint32_t actor_remove_cb(cell_rt *actor, uint32_t id, uint32_t interval)
   
   actor->disrupt = 1;
   
-  if (!JS_IsNull(actor->unneeded)) {
-    JSValue ret = JS_Call(actor->context, actor->unneeded, JS_NULL, 0, NULL);
+  if (!JS_IsNull(actor->unneeded_ref.val)) {
+    JSValue ret = JS_Call(actor->context, actor->unneeded_ref.val, JS_NULL, 0, NULL);
     uncaught_exception(actor->context, ret);
   }
 
@@ -328,11 +325,7 @@ cell_rt *create_actor(void *wota)
 {
   cell_rt *actor = calloc(sizeof(*actor), 1);
   actor->init_wota = wota;
-  actor->idx_buffer = JS_NULL;
-  actor->message_handle = JS_NULL;
-  actor->unneeded = JS_NULL;
-  actor->on_exception = JS_NULL;
-  actor->actor_sym = JS_NULL;
+  /* GCRef fields are registered after JSContext creation in script_startup. */
 
   arrsetcap(actor->letters, 5);
 
@@ -408,7 +401,7 @@ void actor_turn(cell_rt *actor)
     size_t size = blob_length(l.blob_data) / 8; // Convert bits to bytes
     JSValue arg = js_new_blob_stoned_copy(actor->context, (void *)blob_data(l.blob_data), size);
     blob_destroy(l.blob_data);
-    result = JS_Call(actor->context, actor->message_handle, JS_NULL, 1, &arg);
+    result = JS_Call(actor->context, actor->message_handle_ref.val, JS_NULL, 1, &arg);
     uncaught_exception(actor->context, result);
     JS_FreeValue(actor->context, arg);
   } else if (l.type == LETTER_CALLBACK) {

source/suite.c

@@ -1742,9 +1742,12 @@ TEST(has_exception_initially_false) {
   return 1;
 }
 
-TEST(new_error) {
-  JSValue err = JS_NewError(ctx);
-  ASSERT(JS_IsError(ctx, err));
+TEST(throw_and_check_exception) {
+  ASSERT(!JS_HasException(ctx));
+  JS_ThrowTypeError(ctx, "test error");
+  ASSERT(JS_HasException(ctx));
+  JS_GetException(ctx); /* clear it */
+  ASSERT(!JS_HasException(ctx));
   return 1;
 }
 
@@ -2001,496 +2004,11 @@ TEST(wota_encode_blob) {
   return 1;
 }
 
-/* ============================================================================
-   CELL MODULE TESTS - Serialize/Deserialize bytecode
-   ============================================================================ */
-
-TEST(cell_module_compile_basic) {
-  /* Compile simple source to CellModule */
-  const char *source = "1 + 2";
-  CellModule *mod = JS_CompileModule(ctx, source, strlen(source), "<test>");
-  ASSERT_MSG(mod != NULL, "JS_CompileModule returned NULL");
-
-  /* Check module has units */
-  ASSERT_MSG(mod->unit_count > 0, "Module has no units");
-  ASSERT_MSG(mod->units[0].bytecode_len > 0, "Unit has no bytecode");
-
-  cell_module_free(mod);
-  return 1;
-}
-
-TEST(cell_module_write_read) {
-  /* Compile, serialize, deserialize */
-  const char *source = "var x = 10; x * 2";
-  CellModule *mod = JS_CompileModule(ctx, source, strlen(source), "<test>");
-  ASSERT_MSG(mod != NULL, "JS_CompileModule returned NULL");
-
-  /* Serialize */
-  size_t len;
-  uint8_t *buf = cell_module_write(mod, &len);
-  ASSERT_MSG(buf != NULL, "cell_module_write returned NULL");
-  ASSERT_MSG(len > 0, "cell_module_write produced empty buffer");
-
-  /* Deserialize */
-  CellModule *mod2 = cell_module_read(buf, len);
-  free(buf);
-  ASSERT_MSG(mod2 != NULL, "cell_module_read returned NULL");
-
-  /* Verify structure matches */
-  ASSERT_MSG(mod2->unit_count == mod->unit_count, "unit_count mismatch");
-  ASSERT_MSG(mod2->string_count == mod->string_count, "string_count mismatch");
-
-  cell_module_free(mod);
-  cell_module_free(mod2);
-  return 1;
-}
-
-TEST(cell_module_integrate_basic) {
-  /* Compile, then integrate and execute */
-  const char *source = "3 + 4";
-  CellModule *mod = JS_CompileModule(ctx, source, strlen(source), "<test>");
-  ASSERT_MSG(mod != NULL, "JS_CompileModule returned NULL");
-
-  /* Integrate into context */
-  JSValue func = cell_module_integrate(ctx, mod, JS_NULL);
-  if (JS_IsException(func)) {
-    cell_module_free(mod);
-    ASSERT_MSG(0, "cell_module_integrate threw exception");
-  }
-
-  /* Execute */
-  JSValue result = JS_Call(ctx, func, JS_NULL, 0, NULL);
-  JS_FreeValue(ctx, func);
-  cell_module_free(mod);
-
-  if (JS_IsException(result)) {
-    ASSERT_MSG(0, "JS_Call threw exception");
-  }
-
-  ASSERT_INT(result, 7);
-  return 1;
-}
-
-TEST(cell_module_roundtrip_execute) {
-  /* Full round-trip: compile -> write -> read -> integrate -> execute */
-  const char *source = "var a = 5; var b = 3; a * b";
-  CellModule *mod = JS_CompileModule(ctx, source, strlen(source), "<test>");
-  ASSERT_MSG(mod != NULL, "JS_CompileModule returned NULL");
-
-  /* Serialize */
-  size_t len;
-  uint8_t *buf = cell_module_write(mod, &len);
-  cell_module_free(mod);
-  ASSERT_MSG(buf != NULL, "cell_module_write returned NULL");
-
-  /* Deserialize */
-  CellModule *mod2 = cell_module_read(buf, len);
-  free(buf);
-  ASSERT_MSG(mod2 != NULL, "cell_module_read returned NULL");
-
-  /* Integrate and execute */
-  JSValue func = cell_module_integrate(ctx, mod2, JS_NULL);
-  cell_module_free(mod2);
-  if (JS_IsException(func)) {
-    ASSERT_MSG(0, "cell_module_integrate threw exception");
-  }
-
-  JSValue result = JS_Call(ctx, func, JS_NULL, 0, NULL);
-  JS_FreeValue(ctx, func);
-
-  if (JS_IsException(result)) {
-    ASSERT_MSG(0, "JS_Call threw exception");
-  }
-
-  ASSERT_INT(result, 15);
-  return 1;
-}
-
-TEST(cell_module_string_constant) {
-  /* Test string constant handling */
-  const char *source = "'hello' + ' world'";
-  CellModule *mod = JS_CompileModule(ctx, source, strlen(source), "<test>");
-  ASSERT_MSG(mod != NULL, "JS_CompileModule returned NULL");
-
-  /* Verify string table has entries */
-  ASSERT_MSG(mod->string_count > 0, "Module has no strings");
-
-  /* Integrate and execute */
-  JSValue func = cell_module_integrate(ctx, mod, JS_NULL);
-  cell_module_free(mod);
-  if (JS_IsException(func)) {
-    ASSERT_MSG(0, "cell_module_integrate threw exception");
-  }
-
-  JSValue result = JS_Call(ctx, func, JS_NULL, 0, NULL);
-  JS_FreeValue(ctx, func);
-
-  if (JS_IsException(result)) {
-    ASSERT_MSG(0, "JS_Call threw exception");
-  }
-
-  ASSERT_STR(result, "hello world");
-  return 1;
-}
-
-/* ============================================================================
-   ERROR RECOVERY TESTS - Helper macros
-   ============================================================================ */
-
-#define ASSERT_HAS_ERRORS(json_str, min_count) do { \
-  cJSON *_root = cJSON_Parse(json_str); \
-  ASSERT_MSG(_root != NULL, "failed to parse JSON output"); \
-  cJSON *_errs = cJSON_GetObjectItem(_root, "errors"); \
-  if (!_errs || !cJSON_IsArray(_errs) || cJSON_GetArraySize(_errs) < (min_count)) { \
-    printf("[line %d: expected at least %d error(s), got %d] ", __LINE__, (min_count), \
-           _errs && cJSON_IsArray(_errs) ? cJSON_GetArraySize(_errs) : 0); \
-    cJSON_Delete(_root); \
-    return 0; \
-  } \
-  cJSON_Delete(_root); \
-} while(0)
-
-#define ASSERT_NO_ERRORS(json_str) do { \
-  cJSON *_root = cJSON_Parse(json_str); \
-  ASSERT_MSG(_root != NULL, "failed to parse JSON output"); \
-  cJSON *_errs = cJSON_GetObjectItem(_root, "errors"); \
-  if (_errs && cJSON_IsArray(_errs) && cJSON_GetArraySize(_errs) > 0) { \
-    cJSON *_first = cJSON_GetArrayItem(_errs, 0); \
-    const char *_msg = cJSON_GetStringValue(cJSON_GetObjectItem(_first, "message")); \
-    printf("[line %d: expected no errors, got: %s] ", __LINE__, _msg ? _msg : "?"); \
-    cJSON_Delete(_root); \
-    return 0; \
-  } \
-  cJSON_Delete(_root); \
-} while(0)
-
-#define ASSERT_ERROR_MSG_CONTAINS(json_str, substring) do { \
-  cJSON *_root = cJSON_Parse(json_str); \
-  ASSERT_MSG(_root != NULL, "failed to parse JSON output"); \
-  cJSON *_errs = cJSON_GetObjectItem(_root, "errors"); \
-  int _found = 0; \
-  if (_errs && cJSON_IsArray(_errs)) { \
-    cJSON *_e; \
-    cJSON_ArrayForEach(_e, _errs) { \
-      const char *_msg = cJSON_GetStringValue(cJSON_GetObjectItem(_e, "message")); \
-      if (_msg && strstr(_msg, (substring))) { _found = 1; break; } \
-    } \
-  } \
-  if (!_found) { \
-    printf("[line %d: no error containing '%s'] ", __LINE__, (substring)); \
-    cJSON_Delete(_root); \
-    return 0; \
-  } \
-  cJSON_Delete(_root); \
-} while(0)
-
-/* ============================================================================
-   TOKENIZER ERROR TESTS
-   ============================================================================ */
-
-TEST(tokenize_unterminated_string) {
-  const char *src = "var x = \"hello";
-  char *json = JS_Tokenize(ctx, src, strlen(src), "<test>");
-  ASSERT_MSG(json != NULL, "JS_Tokenize returned NULL");
-  ASSERT_ERROR_MSG_CONTAINS(json, "unterminated string");
-  free(json);
-  return 1;
-}
-
-TEST(tokenize_unterminated_template) {
-  const char *src = "var x = `hello";
-  char *json = JS_Tokenize(ctx, src, strlen(src), "<test>");
-  ASSERT_MSG(json != NULL, "JS_Tokenize returned NULL");
-  ASSERT_ERROR_MSG_CONTAINS(json, "unterminated template");
-  free(json);
-  return 1;
-}
-
-TEST(tokenize_unterminated_block_comment) {
-  const char *src = "var x /* comment";
-  char *json = JS_Tokenize(ctx, src, strlen(src), "<test>");
-  ASSERT_MSG(json != NULL, "JS_Tokenize returned NULL");
-  ASSERT_ERROR_MSG_CONTAINS(json, "unterminated block comment");
-  free(json);
-  return 1;
-}
-
-TEST(tokenize_malformed_hex) {
-  const char *src = "var x = 0x";
-  char *json = JS_Tokenize(ctx, src, strlen(src), "<test>");
-  ASSERT_MSG(json != NULL, "JS_Tokenize returned NULL");
-  ASSERT_ERROR_MSG_CONTAINS(json, "malformed hex");
-  free(json);
-  return 1;
-}
-
-TEST(tokenize_malformed_binary) {
-  const char *src = "var x = 0b";
-  char *json = JS_Tokenize(ctx, src, strlen(src), "<test>");
-  ASSERT_MSG(json != NULL, "JS_Tokenize returned NULL");
-  ASSERT_ERROR_MSG_CONTAINS(json, "malformed binary");
-  free(json);
-  return 1;
-}
-
-TEST(tokenize_malformed_exponent) {
-  const char *src = "var x = 1e+";
-  char *json = JS_Tokenize(ctx, src, strlen(src), "<test>");
-  ASSERT_MSG(json != NULL, "JS_Tokenize returned NULL");
-  ASSERT_ERROR_MSG_CONTAINS(json, "no digits after exponent");
-  free(json);
-  return 1;
-}
-
-TEST(tokenize_valid_no_errors) {
-  const char *src = "var x = 42";
-  char *json = JS_Tokenize(ctx, src, strlen(src), "<test>");
-  ASSERT_MSG(json != NULL, "JS_Tokenize returned NULL");
-  ASSERT_NO_ERRORS(json);
-  free(json);
-  return 1;
-}
-
-/* ============================================================================
-   PARSER ERROR TESTS
-   ============================================================================ */
-
-TEST(ast_missing_identifier_after_var) {
-  const char *src = "var = 1";
-  char *json = JS_AST(ctx, src, strlen(src), "<test>");
-  ASSERT_MSG(json != NULL, "JS_AST returned NULL");
-  ASSERT_ERROR_MSG_CONTAINS(json, "expected identifier");
-  free(json);
-  return 1;
-}
-
-TEST(ast_missing_initializer_def) {
-  const char *src = "def x";
-  char *json = JS_AST(ctx, src, strlen(src), "<test>");
-  ASSERT_MSG(json != NULL, "JS_AST returned NULL");
-  ASSERT_ERROR_MSG_CONTAINS(json, "missing initializer");
-  free(json);
-  return 1;
-}
-
-TEST(ast_recovery_continues_after_error) {
-  const char *src = "var = 1; var y = 2";
-  char *json = JS_AST(ctx, src, strlen(src), "<test>");
-  ASSERT_MSG(json != NULL, "JS_AST returned NULL");
-  ASSERT_HAS_ERRORS(json, 1);
-  /* Check that 'y' statement is present in the AST */
-  ASSERT_MSG(strstr(json, "\"y\"") != NULL, "recovery failed: 'y' not in AST");
-  free(json);
-  return 1;
-}
-
-TEST(ast_valid_no_errors) {
-  const char *src = "var x = 1; var y = 2";
-  char *json = JS_AST(ctx, src, strlen(src), "<test>");
-  ASSERT_MSG(json != NULL, "JS_AST returned NULL");
-  ASSERT_NO_ERRORS(json);
-  free(json);
-  return 1;
-}
-
-/* ============================================================================
-   AST SEMANTIC ERROR TESTS
-   ============================================================================ */
-
-TEST(ast_sem_assign_to_const) {
-  const char *src = "def x = 5; x = 3";
-  char *json = JS_AST(ctx, src, strlen(src), "<test>");
-  ASSERT_MSG(json != NULL, "JS_AST returned NULL");
-  ASSERT_ERROR_MSG_CONTAINS(json, "cannot assign to constant");
-  free(json);
-  return 1;
-}
-
-TEST(ast_sem_assign_to_arg) {
-  const char *src = "function(x) { x = 5; }";
-  char *json = JS_AST(ctx, src, strlen(src), "<test>");
-  ASSERT_MSG(json != NULL, "JS_AST returned NULL");
-  ASSERT_ERROR_MSG_CONTAINS(json, "cannot assign to constant");
-  free(json);
-  return 1;
-}
-
-TEST(ast_sem_redeclare_const) {
-  const char *src = "def x = 1; def x = 2";
-  char *json = JS_AST(ctx, src, strlen(src), "<test>");
-  ASSERT_MSG(json != NULL, "JS_AST returned NULL");
-  ASSERT_ERROR_MSG_CONTAINS(json, "cannot redeclare constant");
-  free(json);
-  return 1;
-}
-
-TEST(ast_sem_break_outside_loop) {
-  const char *src = "break";
-  char *json = JS_AST(ctx, src, strlen(src), "<test>");
-  ASSERT_MSG(json != NULL, "JS_AST returned NULL");
-  ASSERT_ERROR_MSG_CONTAINS(json, "outside of loop");
-  free(json);
-  return 1;
-}
-
-TEST(ast_sem_continue_outside_loop) {
-  const char *src = "continue";
-  char *json = JS_AST(ctx, src, strlen(src), "<test>");
-  ASSERT_MSG(json != NULL, "JS_AST returned NULL");
-  ASSERT_ERROR_MSG_CONTAINS(json, "outside of loop");
-  free(json);
-  return 1;
-}
-
-TEST(ast_sem_break_inside_loop_ok) {
-  const char *src = "while (true) { break; }";
-  char *json = JS_AST(ctx, src, strlen(src), "<test>");
-  ASSERT_MSG(json != NULL, "JS_AST returned NULL");
-  ASSERT_NO_ERRORS(json);
-  free(json);
-  return 1;
-}
-
-TEST(ast_sem_increment_const) {
-  const char *src = "def x = 1; x++";
-  char *json = JS_AST(ctx, src, strlen(src), "<test>");
-  ASSERT_MSG(json != NULL, "JS_AST returned NULL");
-  ASSERT_ERROR_MSG_CONTAINS(json, "cannot assign to constant");
-  free(json);
-  return 1;
-}
-
-TEST(ast_sem_shadow_var_ok) {
-  const char *src = "var array = []; array";
-  char *json = JS_AST(ctx, src, strlen(src), "<test>");
-  ASSERT_MSG(json != NULL, "JS_AST returned NULL");
-  ASSERT_NO_ERRORS(json);
-  free(json);
-  return 1;
-}
-
-TEST(ast_sem_var_assign_ok) {
-  const char *src = "var x = 1; x = x + 1";
-  char *json = JS_AST(ctx, src, strlen(src), "<test>");
-  ASSERT_MSG(json != NULL, "JS_AST returned NULL");
-  ASSERT_NO_ERRORS(json);
-  free(json);
-  return 1;
-}
-
-TEST(ast_sem_nested_function_scope) {
-  const char *src = "var x = 1; function f(x) { return x + 1; }";
-  char *json = JS_AST(ctx, src, strlen(src), "<test>");
-  ASSERT_MSG(json != NULL, "JS_AST returned NULL");
-  ASSERT_NO_ERRORS(json);
-  free(json);
-  return 1;
-}
-
-/* ============================================================================
-   CODEGEN/SEMANTIC ERROR TESTS
-   ============================================================================ */
-
-TEST(mach_assign_to_const) {
-  const char *src = "def x = 1; x = 2";
-  char *ast_json = JS_AST(ctx, src, strlen(src), "<test>");
-  ASSERT_MSG(ast_json != NULL, "JS_AST returned NULL");
-  char *mach_json = JS_Mach(ctx, ast_json);
-  free(ast_json);
-  ASSERT_MSG(mach_json != NULL, "JS_Mach returned NULL");
-  ASSERT_ERROR_MSG_CONTAINS(mach_json, "cannot assign to constant");
-  free(mach_json);
-  return 1;
-}
-
-TEST(mach_assign_to_arg) {
-  const char *src = "function f(x) { x = 5 }";
-  char *ast_json = JS_AST(ctx, src, strlen(src), "<test>");
-  ASSERT_MSG(ast_json != NULL, "JS_AST returned NULL");
-  char *mach_json = JS_Mach(ctx, ast_json);
-  free(ast_json);
-  ASSERT_MSG(mach_json != NULL, "JS_Mach returned NULL");
-  ASSERT_ERROR_MSG_CONTAINS(mach_json, "cannot assign to function argument");
-  free(mach_json);
-  return 1;
-}
-
-TEST(mach_redeclare_const) {
-  const char *src = "def x = 1; def x = 2";
-  char *ast_json = JS_AST(ctx, src, strlen(src), "<test>");
-  ASSERT_MSG(ast_json != NULL, "JS_AST returned NULL");
-  char *mach_json = JS_Mach(ctx, ast_json);
-  free(ast_json);
-  ASSERT_MSG(mach_json != NULL, "JS_Mach returned NULL");
-  ASSERT_ERROR_MSG_CONTAINS(mach_json, "cannot redeclare constant");
-  free(mach_json);
-  return 1;
-}
-
-TEST(mach_break_outside_loop) {
-  const char *src = "break";
-  char *ast_json = JS_AST(ctx, src, strlen(src), "<test>");
-  ASSERT_MSG(ast_json != NULL, "JS_AST returned NULL");
-  char *mach_json = JS_Mach(ctx, ast_json);
-  free(ast_json);
-  ASSERT_MSG(mach_json != NULL, "JS_Mach returned NULL");
-  ASSERT_ERROR_MSG_CONTAINS(mach_json, "outside of loop");
-  free(mach_json);
-  return 1;
-}
-
-TEST(mach_continue_outside_loop) {
-  const char *src = "continue";
-  char *ast_json = JS_AST(ctx, src, strlen(src), "<test>");
-  ASSERT_MSG(ast_json != NULL, "JS_AST returned NULL");
-  char *mach_json = JS_Mach(ctx, ast_json);
-  free(ast_json);
-  ASSERT_MSG(mach_json != NULL, "JS_Mach returned NULL");
-  ASSERT_ERROR_MSG_CONTAINS(mach_json, "outside of loop");
-  free(mach_json);
-  return 1;
-}
-
-TEST(mach_assign_unbound_var) {
-  const char *src = "x = 42";
-  char *ast_json = JS_AST(ctx, src, strlen(src), "<test>");
-  ASSERT_MSG(ast_json != NULL, "JS_AST returned NULL");
-  char *mach_json = JS_Mach(ctx, ast_json);
-  free(ast_json);
-  ASSERT_MSG(mach_json != NULL, "JS_Mach returned NULL");
-  ASSERT_ERROR_MSG_CONTAINS(mach_json, "not declared");
-  free(mach_json);
-  return 1;
-}
-
-TEST(mach_shadow_is_ok) {
-  const char *src = "var array = []; array";
-  char *ast_json = JS_AST(ctx, src, strlen(src), "<test>");
-  ASSERT_MSG(ast_json != NULL, "JS_AST returned NULL");
-  char *mach_json = JS_Mach(ctx, ast_json);
-  free(ast_json);
-  ASSERT_MSG(mach_json != NULL, "JS_Mach returned NULL");
-  ASSERT_NO_ERRORS(mach_json);
-  free(mach_json);
-  return 1;
-}
-
-TEST(mach_valid_no_errors) {
-  const char *src = "var x = 1; x = x + 1";
-  char *ast_json = JS_AST(ctx, src, strlen(src), "<test>");
-  ASSERT_MSG(ast_json != NULL, "JS_AST returned NULL");
-  char *mach_json = JS_Mach(ctx, ast_json);
-  free(ast_json);
-  ASSERT_MSG(mach_json != NULL, "JS_Mach returned NULL");
-  ASSERT_NO_ERRORS(mach_json);
-  free(mach_json);
-  return 1;
-}
-
 /* ============================================================================
    MAIN TEST RUNNER
    ============================================================================ */
 
+
 int run_c_test_suite(JSContext *ctx)
 {
   printf("\n=== Cell Runtime C Test Suite ===\n\n");
@@ -2664,7 +2182,7 @@ int run_c_test_suite(JSContext *ctx)
 
   printf("\nExceptions/Errors:\n");
   RUN_TEST(has_exception_initially_false);
-  RUN_TEST(new_error);
+  RUN_TEST(throw_and_check_exception);
 
   printf("\nFloat Edge Cases:\n");
   RUN_TEST(float_small_positive);
@@ -2696,49 +2214,6 @@ int run_c_test_suite(JSContext *ctx)
   RUN_TEST(wota_encode_nested_array);
   RUN_TEST(wota_encode_blob);
 
-  // CellModule tests
-  RUN_TEST(cell_module_compile_basic);
-  RUN_TEST(cell_module_write_read);
-  RUN_TEST(cell_module_integrate_basic);
-  RUN_TEST(cell_module_roundtrip_execute);
-  RUN_TEST(cell_module_string_constant);
-
-  printf("\nTokenizer Errors:\n");
-  RUN_TEST(tokenize_unterminated_string);
-  RUN_TEST(tokenize_unterminated_template);
-  RUN_TEST(tokenize_unterminated_block_comment);
-  RUN_TEST(tokenize_malformed_hex);
-  RUN_TEST(tokenize_malformed_binary);
-  RUN_TEST(tokenize_malformed_exponent);
-  RUN_TEST(tokenize_valid_no_errors);
-
-  printf("\nParser Errors:\n");
-  RUN_TEST(ast_missing_identifier_after_var);
-  RUN_TEST(ast_missing_initializer_def);
-  RUN_TEST(ast_recovery_continues_after_error);
-  RUN_TEST(ast_valid_no_errors);
-
-  printf("\nAST Semantic Errors:\n");
-  RUN_TEST(ast_sem_assign_to_const);
-  RUN_TEST(ast_sem_assign_to_arg);
-  RUN_TEST(ast_sem_redeclare_const);
-  RUN_TEST(ast_sem_break_outside_loop);
-  RUN_TEST(ast_sem_continue_outside_loop);
-  RUN_TEST(ast_sem_break_inside_loop_ok);
-  RUN_TEST(ast_sem_increment_const);
-  RUN_TEST(ast_sem_shadow_var_ok);
-  RUN_TEST(ast_sem_var_assign_ok);
-  RUN_TEST(ast_sem_nested_function_scope);
-
-  printf("\nCodegen Errors:\n");
-  RUN_TEST(mach_assign_to_const);
-  RUN_TEST(mach_assign_to_arg);
-  RUN_TEST(mach_redeclare_const);
-  RUN_TEST(mach_break_outside_loop);
-  RUN_TEST(mach_continue_outside_loop);
-  RUN_TEST(mach_assign_unbound_var);
-  RUN_TEST(mach_shadow_is_ok);
-  RUN_TEST(mach_valid_no_errors);
 
   printf("\n=================================\n");
   printf("Results: %d passed, %d failed\n", tests_passed, tests_failed);

syntax_suite.ce

@@ -0,0 +1,641 @@
+// Syntax suite: covers every syntactic feature of cell script
+// Run: ./cell --mach-run syntax_suite.ce
+
+var passed = 0
+var failed = 0
+var error_names = []
+var error_reasons = []
+var fail_msg = ""
+
+var _i = 0
+for (_i = 0; _i < 100; _i++) {
+  error_names[] = null
+  error_reasons[] = null
+}
+
+var fail = function(msg) {
+  fail_msg = msg
+  disrupt
+}
+
+var assert_eq = function(actual, expected, msg) {
+  if (actual != expected) fail(msg + " (got=" + text(actual) + " expected=" + text(expected) + ")")
+}
+
+var run = function(name, fn) {
+  fail_msg = ""
+  fn()
+  passed = passed + 1
+} disruption {
+  error_names[failed] = name
+  error_reasons[failed] = fail_msg == "" ? "disruption" : fail_msg
+  failed = failed + 1
+}
+
+var should_disrupt = function(fn) {
+  var caught = false
+  var wrapper = function() {
+    fn()
+  } disruption {
+    caught = true
+  }
+  wrapper()
+  return caught
+}
+
+// === LITERALS ===
+
+run("number literals", function() {
+  assert_eq(42, 42, "integer")
+  assert_eq(3.14 > 3, true, "float")
+  assert_eq(-5, -5, "negative")
+  assert_eq(0, 0, "zero")
+  assert_eq(1e3, 1000, "scientific")
+})
+
+run("string literals", function() {
+  assert_eq("hello", "hello", "double quote")
+  assert_eq("", "", "empty string")
+  assert_eq("line1\nline2" != "line1line2", true, "escape sequence")
+})
+
+run("template literals", function() {
+  var x = "world"
+  assert_eq(`hello ${x}`, "hello world", "interpolation")
+  assert_eq(`${1 + 2}`, "3", "expression interpolation")
+  assert_eq(`plain`, "plain", "no interpolation")
+})
+
+run("boolean literals", function() {
+  assert_eq(true, true, "true")
+  assert_eq(false, false, "false")
+})
+
+run("null literal", function() {
+  assert_eq(null, null, "null")
+})
+
+run("array literal", function() {
+  var a = [1, 2, 3]
+  assert_eq(length(a), 3, "array length")
+  assert_eq(a[0], 1, "first element")
+  var e = []
+  assert_eq(length(e), 0, "empty array")
+})
+
+run("object literal", function() {
+  var o = {a: 1, b: "two"}
+  assert_eq(o.a, 1, "object prop")
+  var e = {}
+  assert_eq(e.x, null, "empty object missing prop")
+})
+
+run("regex literal", function() {
+  var r = /\d+/
+  var result = extract("abc123", r)
+  assert_eq(result[0], "123", "regex match")
+  var ri = /hello/i
+  var result2 = extract("Hello", ri)
+  assert_eq(result2[0], "Hello", "regex flags")
+})
+
+// === DECLARATIONS ===
+
+run("var declaration", function() {
+  var x = 5
+  assert_eq(x, 5, "var init")
+})
+
+run("var multiple declaration", function() {
+  var a = 1, b = 2, c = 3
+  assert_eq(a + b + c, 6, "multi var")
+})
+
+run("def declaration", function() {
+  def x = 42
+  assert_eq(x, 42, "def const")
+})
+
+// === ARITHMETIC OPERATORS ===
+
+run("arithmetic operators", function() {
+  assert_eq(2 + 3, 5, "add")
+  assert_eq(5 - 3, 2, "sub")
+  assert_eq(3 * 4, 12, "mul")
+  assert_eq(12 / 4, 3, "div")
+  assert_eq(10 % 3, 1, "mod")
+  assert_eq(2 ** 3, 8, "exp")
+})
+
+// === COMPARISON OPERATORS ===
+
+run("comparison operators", function() {
+  assert_eq(5 == 5, true, "eq")
+  assert_eq(5 != 6, true, "neq")
+  assert_eq(3 < 5, true, "lt")
+  assert_eq(5 > 3, true, "gt")
+  assert_eq(3 <= 3, true, "lte")
+  assert_eq(5 >= 5, true, "gte")
+})
+
+// === LOGICAL OPERATORS ===
+
+run("logical operators", function() {
+  assert_eq(true && true, true, "and")
+  assert_eq(true && false, false, "and false")
+  assert_eq(false || true, true, "or")
+  assert_eq(false || false, false, "or false")
+  assert_eq(!true, false, "not")
+  assert_eq(!false, true, "not false")
+})
+
+run("short circuit", function() {
+  var called = false
+  var fn = function() { called = true; return true }
+  var r = false && fn()
+  assert_eq(called, false, "and short circuit")
+  r = true || fn()
+  assert_eq(called, false, "or short circuit")
+})
+
+// === BITWISE OPERATORS ===
+
+run("bitwise operators", function() {
+  assert_eq(5 & 3, 1, "and")
+  assert_eq(5 | 3, 7, "or")
+  assert_eq(5 ^ 3, 6, "xor")
+  assert_eq(~0, -1, "not")
+  assert_eq(1 << 3, 8, "lshift")
+  assert_eq(8 >> 3, 1, "rshift")
+  assert_eq(-1 >>> 1, 2147483647, "unsigned rshift")
+})
+
+// === UNARY OPERATORS ===
+
+run("unary operators", function() {
+  assert_eq(+5, 5, "unary plus")
+  assert_eq(-5, -5, "unary minus")
+  assert_eq(-(-5), 5, "double negate")
+})
+
+run("increment decrement", function() {
+  var x = 5
+  assert_eq(x++, 5, "postfix inc returns old")
+  assert_eq(x, 6, "postfix inc side effect")
+  x = 5
+  assert_eq(++x, 6, "prefix inc returns new")
+  x = 5
+  assert_eq(x--, 5, "postfix dec returns old")
+  assert_eq(x, 4, "postfix dec side effect")
+  x = 5
+  assert_eq(--x, 4, "prefix dec returns new")
+})
+
+// === COMPOUND ASSIGNMENT ===
+
+run("compound assignment", function() {
+  var x = 10
+  x += 3; assert_eq(x, 13, "+=")
+  x -= 3; assert_eq(x, 10, "-=")
+  x *= 2; assert_eq(x, 20, "*=")
+  x /= 4; assert_eq(x, 5, "/=")
+  x %= 3; assert_eq(x, 2, "%=")
+})
+
+// === TERNARY OPERATOR ===
+
+run("ternary operator", function() {
+  var a = true ? 1 : 2
+  assert_eq(a, 1, "ternary true")
+  var b = false ? 1 : 2
+  assert_eq(b, 2, "ternary false")
+  var c = true ? (false ? 1 : 2) : 3
+  assert_eq(c, 2, "ternary nested")
+})
+
+// === COMMA OPERATOR ===
+
+run("comma operator", function() {
+  var x = (1, 2, 3)
+  assert_eq(x, 3, "comma returns last")
+})
+
+// === IN OPERATOR ===
+
+run("in operator", function() {
+  var o = {a: 1}
+  assert_eq("a" in o, true, "key exists")
+  assert_eq("b" in o, false, "key missing")
+})
+
+// === DELETE OPERATOR ===
+
+run("delete operator", function() {
+  var o = {a: 1, b: 2}
+  delete o.a
+  assert_eq("a" in o, false, "delete removes key")
+  assert_eq(o.b, 2, "delete leaves others")
+})
+
+// === PROPERTY ACCESS ===
+
+run("dot access", function() {
+  var o = {x: 10}
+  assert_eq(o.x, 10, "dot read")
+  o.x = 20
+  assert_eq(o.x, 20, "dot write")
+})
+
+run("bracket access", function() {
+  var o = {x: 10}
+  assert_eq(o["x"], 10, "bracket read")
+  var key = "x"
+  assert_eq(o[key], 10, "computed bracket")
+  o["y"] = 20
+  assert_eq(o.y, 20, "bracket write")
+})
+
+run("object-as-key", function() {
+  var k = {}
+  var o = {}
+  o[k] = 42
+  assert_eq(o[k], 42, "object key set/get")
+  assert_eq(o[{}], null, "new object is different key")
+  assert_eq(k in o, true, "object key in")
+  delete o[k]
+  assert_eq(k in o, false, "object key delete")
+})
+
+run("chained access", function() {
+  var d = {a: {b: [1, {c: 99}]}}
+  assert_eq(d.a.b[1].c, 99, "mixed chain")
+})
+
+// === ARRAY PUSH/POP SYNTAX ===
+
+run("array push pop", function() {
+  var a = [1, 2]
+  a[] = 3
+  assert_eq(length(a), 3, "push length")
+  assert_eq(a[2], 3, "push value")
+  var v = a[]
+  assert_eq(v, 3, "pop value")
+  assert_eq(length(a), 2, "pop length")
+})
+
+// === CONTROL FLOW: IF/ELSE ===
+
+run("if else", function() {
+  var x = 0
+  if (true) x = 1
+  assert_eq(x, 1, "if true")
+  if (false) x = 2 else x = 3
+  assert_eq(x, 3, "if else")
+  if (false) x = 4
+  else if (true) x = 5
+  else x = 6
+  assert_eq(x, 5, "else if")
+})
+
+// === CONTROL FLOW: WHILE ===
+
+run("while loop", function() {
+  var i = 0
+  while (i < 5) i++
+  assert_eq(i, 5, "while basic")
+})
+
+run("while break continue", function() {
+  var i = 0
+  while (true) {
+    if (i >= 3) break
+    i++
+  }
+  assert_eq(i, 3, "while break")
+  var sum = 0
+  i = 0
+  while (i < 5) {
+    i++
+    if (i % 2 == 0) continue
+    sum += i
+  }
+  assert_eq(sum, 9, "while continue")
+})
+
+// === CONTROL FLOW: FOR ===
+
+run("for loop", function() {
+  var sum = 0
+  var i = 0
+  for (i = 0; i < 5; i++) sum += i
+  assert_eq(sum, 10, "for basic")
+})
+
+run("for break continue", function() {
+  var sum = 0
+  var i = 0
+  for (i = 0; i < 10; i++) {
+    if (i == 5) break
+    sum += i
+  }
+  assert_eq(sum, 10, "for break")
+  sum = 0
+  for (i = 0; i < 5; i++) {
+    if (i % 2 == 0) continue
+    sum += i
+  }
+  assert_eq(sum, 4, "for continue")
+})
+
+run("nested for", function() {
+  var sum = 0
+  var i = 0, j = 0
+  for (i = 0; i < 3; i++)
+    for (j = 0; j < 3; j++)
+      sum++
+  assert_eq(sum, 9, "nested for")
+})
+
+// === FUNCTIONS ===
+
+run("function expression", function() {
+  var fn = function(a, b) { return a + b }
+  assert_eq(fn(2, 3), 5, "basic call")
+})
+
+run("arrow function", function() {
+  var double = x => x * 2
+  assert_eq(double(5), 10, "arrow single param")
+  var add = (a, b) => a + b
+  assert_eq(add(2, 3), 5, "arrow multi param")
+  var block = x => {
+    var y = x * 2
+    return y + 1
+  }
+  assert_eq(block(5), 11, "arrow block body")
+})
+
+run("function no return", function() {
+  var fn = function() { var x = 1 }
+  assert_eq(fn(), null, "no return gives null")
+})
+
+run("function early return", function() {
+  var fn = function() { return 1; return 2 }
+  assert_eq(fn(), 1, "early return")
+})
+
+run("extra and missing args", function() {
+  var fn = function(a, b) { return a + b }
+  assert_eq(fn(1, 2, 3), 3, "extra args ignored")
+  var fn2 = function(a, b) { return a }
+  assert_eq(fn2(1), 1, "missing args ok")
+})
+
+run("iife", function() {
+  var r = (function(x) { return x * 2 })(21)
+  assert_eq(r, 42, "immediately invoked")
+})
+
+// === CLOSURES ===
+
+run("closure", function() {
+  var make = function(x) {
+    return function(y) { return x + y }
+  }
+  var add5 = make(5)
+  assert_eq(add5(3), 8, "closure captures")
+})
+
+run("closure mutation", function() {
+  var counter = function() {
+    var n = 0
+    return function() { n = n + 1; return n }
+  }
+  var c = counter()
+  assert_eq(c(), 1, "first")
+  assert_eq(c(), 2, "second")
+})
+
+// === RECURSION ===
+
+run("recursion", function() {
+  var fact = function(n) {
+    if (n <= 1) return 1
+    return n * fact(n - 1)
+  }
+  assert_eq(fact(5), 120, "factorial")
+})
+
+// === THIS BINDING ===
+
+run("this binding", function() {
+  var obj = {
+    val: 10,
+    get: function() { return this.val }
+  }
+  assert_eq(obj.get(), 10, "method this")
+})
+
+// === DISRUPTION ===
+
+run("disrupt keyword", function() {
+  assert_eq(should_disrupt(function() { disrupt }), true, "bare disrupt")
+})
+
+run("disruption handler", function() {
+  var x = 0
+  var fn = function() { x = 1 } disruption { x = 2 }
+  fn()
+  assert_eq(x, 1, "no disruption path")
+  var fn2 = function() { disrupt } disruption { x = 3 }
+  fn2()
+  assert_eq(x, 3, "disruption caught")
+})
+
+run("disruption re-raise", function() {
+  var outer_caught = false
+  var outer = function() {
+    var inner = function() { disrupt } disruption { disrupt }
+    inner()
+  } disruption {
+    outer_caught = true
+  }
+  outer()
+  assert_eq(outer_caught, true, "re-raise propagates")
+})
+
+// === PROTOTYPAL INHERITANCE ===
+
+run("meme and proto", function() {
+  var parent = {x: 10}
+  var child = meme(parent)
+  assert_eq(child.x, 10, "inherited prop")
+  assert_eq(proto(child), parent, "proto returns parent")
+  child.x = 20
+  assert_eq(parent.x, 10, "override does not mutate parent")
+})
+
+run("meme with mixins", function() {
+  var p = {a: 1}
+  var m1 = {b: 2}
+  var m2 = {c: 3}
+  var child = meme(p, [m1, m2])
+  assert_eq(child.a, 1, "parent prop")
+  assert_eq(child.b, 2, "mixin1")
+  assert_eq(child.c, 3, "mixin2")
+})
+
+// === STONE (FREEZE) ===
+
+run("stone", function() {
+  var o = {x: 1}
+  assert_eq(is_stone(o), false, "not frozen")
+  stone(o)
+  assert_eq(is_stone(o), true, "frozen")
+  assert_eq(should_disrupt(function() { o.x = 2 }), true, "write disrupts")
+})
+
+// === FUNCTION PROXY ===
+
+run("function proxy", function() {
+  var proxy = function(name, args) {
+    return `${name}:${length(args)}`
+  }
+  assert_eq(proxy.hello(), "hello:0", "proxy dot call")
+  assert_eq(proxy.add(1, 2), "add:2", "proxy with args")
+  assert_eq(proxy["method"](), "method:0", "proxy bracket call")
+  var m = "dynamic"
+  assert_eq(proxy[m](), "dynamic:0", "proxy computed name")
+})
+
+run("non-proxy function prop access disrupts", function() {
+  var fn = function() { return 1 }
+  assert_eq(should_disrupt(function() { var x = fn.foo }), true, "prop read disrupts")
+  assert_eq(should_disrupt(function() { fn.foo = 1 }), true, "prop write disrupts")
+})
+
+// === TYPE CHECKING ===
+
+run("is_* functions", function() {
+  assert_eq(is_number(42), true, "is_number")
+  assert_eq(is_text("hi"), true, "is_text")
+  assert_eq(is_logical(true), true, "is_logical")
+  assert_eq(is_object({}), true, "is_object")
+  assert_eq(is_array([]), true, "is_array")
+  assert_eq(is_function(function(){}), true, "is_function")
+  assert_eq(is_null(null), true, "is_null")
+  assert_eq(is_object([]), false, "array not object")
+  assert_eq(is_array({}), false, "object not array")
+})
+
+// === TRUTHINESS / FALSINESS ===
+
+run("falsy values", function() {
+  if (false) fail("false")
+  if (0) fail("0")
+  if ("") fail("empty string")
+  if (null) fail("null")
+  assert_eq(true, true, "all falsy passed")
+})
+
+run("truthy values", function() {
+  if (!1) fail("1")
+  if (!"hi") fail("string")
+  if (!{}) fail("object")
+  if (![]) fail("array")
+  if (!true) fail("true")
+  assert_eq(true, true, "all truthy passed")
+})
+
+// === VARIABLE SHADOWING ===
+
+run("variable shadowing", function() {
+  var x = 10
+  var fn = function() {
+    var x = 20
+    return x
+  }
+  assert_eq(fn(), 20, "inner shadows")
+  assert_eq(x, 10, "outer unchanged")
+})
+
+// === OPERATOR PRECEDENCE ===
+
+run("precedence", function() {
+  assert_eq(2 + 3 * 4, 14, "mul before add")
+  assert_eq((2 + 3) * 4, 20, "parens override")
+  assert_eq(-2 * 3, -6, "unary before mul")
+})
+
+// === CURRYING / HIGHER-ORDER ===
+
+run("curried function", function() {
+  var f = function(a) {
+    return function(b) {
+      return function(c) { return a + b + c }
+    }
+  }
+  assert_eq(f(1)(2)(3), 6, "triple curry")
+})
+
+// === SELF-REFERENCING STRUCTURES ===
+
+run("self-referencing object", function() {
+  var o = {name: "root"}
+  o.self = o
+  assert_eq(o.self.self.name, "root", "cycle access")
+})
+
+// === IDENTIFIER ? AND ! ===
+
+run("question mark in identifier", function() {
+  var nil? = (x) => x == null
+  assert_eq(nil?(null), true, "nil? null")
+  assert_eq(nil?(42), false, "nil? 42")
+})
+
+run("bang in identifier", function() {
+  var set! = (x) => x + 1
+  assert_eq(set!(5), 6, "set! call")
+})
+
+run("question mark mid identifier", function() {
+  var is?valid = (x) => x > 0
+  assert_eq(is?valid(3), true, "is?valid true")
+  assert_eq(is?valid(-1), false, "is?valid false")
+})
+
+run("bang mid identifier", function() {
+  var do!stuff = () => 42
+  assert_eq(do!stuff(), 42, "do!stuff call")
+})
+
+run("ternary after question ident", function() {
+  var nil? = (x) => x == null
+  var a = nil?(null) ? "yes" : "no"
+  assert_eq(a, "yes", "ternary true branch")
+  var b = nil?(42) ? "yes" : "no"
+  assert_eq(b, "no", "ternary false branch")
+})
+
+run("bang not confused with logical not", function() {
+  assert_eq(!true, false, "logical not true")
+  assert_eq(!false, true, "logical not false")
+})
+
+run("inequality not confused with bang ident", function() {
+  assert_eq(1 != 2, true, "inequality true")
+  assert_eq(1 != 1, false, "inequality false")
+})
+
+// === SUMMARY ===
+
+print(text(passed) + " passed, " + text(failed) + " failed out of " + text(passed + failed))
+var _j = 0
+if (failed > 0) {
+  print("")
+  for (_j = 0; _j < failed; _j++) {
+    print("  FAIL " + error_names[_j] + ": " + error_reasons[_j])
+  }
+}

tests/demo.ce

@@ -0,0 +1,39 @@
+function safe_add(a, b) {
+  return a + b
+} disruption {
+  print("disruption caught in safe_add")
+}
+
+function inner() {
+  disrupt
+}
+
+function outer() {
+  inner()
+} disruption {
+  print("disruption caught in outer — from inner()")
+}
+
+// Test 1: explicit disrupt with handler
+function test_explicit() {
+  disrupt
+} disruption {
+  print("test 1: explicit disrupt handled")
+}
+test_explicit()
+
+// Test 2: type error disrupt (number + function)
+safe_add(1, print)
+
+// Test 3: unwinding — inner disrupts, outer catches
+outer()
+
+// Test 4: disrupt from inside disruption clause
+function test_nested() {
+  disrupt
+} disruption {
+  print("test 4: first disruption")
+}
+test_nested()
+
+print("done")

tokenize.ce

@@ -0,0 +1,6 @@
+var fd = use("fd")
+var tokenize = use("tokenize")
+var filename = args[0]
+var src = text(fd.slurp(filename))
+var result = tokenize(src, filename)
+print(json.encode({filename: result.filename, tokens: result.tokens}))

tokenize.cm

@@ -0,0 +1,516 @@
+var tokenize = function(src, filename) {
+  var len = length(src)
+  var cp = []
+  var _i = 0
+  while (_i < len) {
+    push(cp, codepoint(src[_i]))
+    _i = _i + 1
+  }
+
+  var pos = 0
+  var row = 0
+  var col = 0
+  var tokens = []
+
+  // Codepoint constants
+  def CP_LF = 10
+  def CP_CR = 13
+  def CP_TAB = 9
+  def CP_SPACE = 32
+  def CP_BANG = 33
+  def CP_DQUOTE = 34
+  def CP_HASH = 35
+  def CP_DOLLAR = 36
+  def CP_PERCENT = 37
+  def CP_AMP = 38
+  def CP_SQUOTE = 39
+  def CP_LPAREN = 40
+  def CP_RPAREN = 41
+  def CP_STAR = 42
+  def CP_PLUS = 43
+  def CP_COMMA = 44
+  def CP_MINUS = 45
+  def CP_DOT = 46
+  def CP_SLASH = 47
+  def CP_0 = 48
+  def CP_1 = 49
+  def CP_7 = 55
+  def CP_9 = 57
+  def CP_COLON = 58
+  def CP_SEMI = 59
+  def CP_LT = 60
+  def CP_EQ = 61
+  def CP_GT = 62
+  def CP_QMARK = 63
+  def CP_AT = 64
+  def CP_A = 65
+  def CP_B = 66
+  def CP_E = 69
+  def CP_F = 70
+  def CP_O = 79
+  def CP_X = 88
+  def CP_Z = 90
+  def CP_LBRACKET = 91
+  def CP_BSLASH = 92
+  def CP_RBRACKET = 93
+  def CP_CARET = 94
+  def CP_UNDERSCORE = 95
+  def CP_BACKTICK = 96
+  def CP_a = 97
+  def CP_b = 98
+  def CP_e = 101
+  def CP_f = 102
+  def CP_n = 110
+  def CP_o = 111
+  def CP_r = 114
+  def CP_t = 116
+  def CP_u = 117
+  def CP_x = 120
+  def CP_z = 122
+  def CP_LBRACE = 123
+  def CP_PIPE = 124
+  def CP_RBRACE = 125
+  def CP_TILDE = 126
+
+  // Keywords lookup
+  var keywords = {
+    if: "if", in: "in", do: "do", go: "go",
+    var: "var", def: "def", for: "for",
+    else: "else", this: "this", null: "null", true: "true",
+    false: "false", while: "while", break: "break",
+    return: "return", delete: "delete",
+    disrupt: "disrupt", function: "function", continue: "continue",
+    disruption: "disruption"
+  }
+
+  var pk = function() {
+    if (pos >= len) return -1
+    return cp[pos]
+  }
+
+  var pk_at = function(n) {
+    var idx = pos + n
+    if (idx >= len) return -1
+    return cp[idx]
+  }
+
+  var adv = function() {
+    var c = cp[pos]
+    pos = pos + 1
+    if (c == CP_LF) {
+      row = row + 1
+      col = 0
+    } else {
+      col = col + 1
+    }
+    return c
+  }
+
+  var is_digit = function(c) {
+    return c >= CP_0 && c <= CP_9
+  }
+
+  var is_hex = function(c) {
+    return (c >= CP_0 && c <= CP_9) || (c >= CP_a && c <= CP_f) || (c >= CP_A && c <= CP_F)
+  }
+
+  var hex_val = function(c) {
+    if (c >= CP_0 && c <= CP_9) return c - CP_0
+    if (c >= CP_a && c <= CP_f) return c - CP_a + 10
+    if (c >= CP_A && c <= CP_F) return c - CP_A + 10
+    return 0
+  }
+
+  var read_unicode_escape = function() {
+    var cp_val = 0
+    var hi = 0
+    while (hi < 4 && pos < len && is_hex(pk())) {
+      cp_val = cp_val * 16 + hex_val(adv())
+      hi = hi + 1
+    }
+    return character(cp_val)
+  }
+
+  var is_alpha = function(c) {
+    return (c >= CP_a && c <= CP_z) || (c >= CP_A && c <= CP_Z)
+  }
+
+  var is_alnum = function(c) {
+    return is_alpha(c) || is_digit(c)
+  }
+
+  var is_ident_start = function(c) {
+    return is_alpha(c) || c == CP_UNDERSCORE || c == CP_DOLLAR
+  }
+
+  var is_ident_char = function(c) {
+    return is_alnum(c) || c == CP_UNDERSCORE || c == CP_DOLLAR || c == CP_QMARK || c == CP_BANG
+  }
+
+  var substr = function(start, end) {
+    var s = ""
+    var i = start
+    while (i < end) {
+      s = s + character(cp[i])
+      i = i + 1
+    }
+    return s
+  }
+
+  var read_string = function(quote_cp) {
+    var start = pos
+    var start_row = row
+    var start_col = col
+    var value = ""
+    var esc = 0
+    adv() // skip opening quote
+    while (pos < len && pk() != quote_cp) {
+      if (pk() == CP_BSLASH) {
+        adv()
+        esc = adv()
+        if (esc == CP_n) { value = value + "\n" }
+        else if (esc == CP_t) { value = value + "\t" }
+        else if (esc == CP_r) { value = value + "\r" }
+        else if (esc == CP_BSLASH) { value = value + "\\" }
+        else if (esc == CP_SQUOTE) { value = value + "'" }
+        else if (esc == CP_DQUOTE) { value = value + "\"" }
+        else if (esc == CP_0) { value = value + character(0) }
+        else if (esc == CP_BACKTICK) { value = value + "`" }
+        else if (esc == CP_u) { value = value + read_unicode_escape() }
+        else { value = value + character(esc) }
+      } else {
+        value = value + character(adv())
+      }
+    }
+    if (pos < len) adv() // skip closing quote
+    push(tokens, {
+      kind: "text", at: start,
+      from_row: start_row, from_column: start_col,
+      to_row: row, to_column: col,
+      value: value
+    })
+  }
+
+  var read_template = function() {
+    var start = pos
+    var start_row = row
+    var start_col = col
+    var value = ""
+    var depth = 0
+    var tc = 0
+    var q = 0
+    adv() // skip opening backtick
+    while (pos < len && pk() != CP_BACKTICK) {
+      if (pk() == CP_BSLASH && pos + 1 < len) {
+        value = value + character(adv())
+        value = value + character(adv())
+      } else if (pk() == CP_DOLLAR && pos + 1 < len && pk_at(1) == CP_LBRACE) {
+        value = value + character(adv()) // $
+        value = value + character(adv()) // {
+        depth = 1
+        while (pos < len && depth > 0) {
+          tc = pk()
+          if (tc == CP_LBRACE) { depth = depth + 1; value = value + character(adv()) }
+          else if (tc == CP_RBRACE) {
+            depth = depth - 1
+            if (depth > 0) { value = value + character(adv()) }
+            else { value = value + character(adv()) }
+          }
+          else if (tc == CP_SQUOTE || tc == CP_DQUOTE || tc == CP_BACKTICK) {
+            q = adv()
+            value = value + character(q)
+            while (pos < len && pk() != q) {
+              if (pk() == CP_BSLASH && pos + 1 < len) {
+                value = value + character(adv())
+              }
+              value = value + character(adv())
+            }
+            if (pos < len) { value = value + character(adv()) }
+          } else { value = value + character(adv()) }
+        }
+      } else {
+        value = value + character(adv())
+      }
+    }
+    if (pos < len) adv() // skip closing backtick
+    push(tokens, {
+      kind: "text", at: start,
+      from_row: start_row, from_column: start_col,
+      to_row: row, to_column: col,
+      value: value
+    })
+  }
+
+  var read_number = function() {
+    var start = pos
+    var start_row = row
+    var start_col = col
+    var raw = ""
+    if (pk() == CP_0 && (pk_at(1) == CP_x || pk_at(1) == CP_X)) {
+      adv(); adv()
+      while (pos < len && (is_hex(pk()) || pk() == CP_UNDERSCORE)) adv()
+    } else if (pk() == CP_0 && (pk_at(1) == CP_b || pk_at(1) == CP_B)) {
+      adv(); adv()
+      while (pos < len && (pk() == CP_0 || pk() == CP_1 || pk() == CP_UNDERSCORE)) adv()
+    } else if (pk() == CP_0 && (pk_at(1) == CP_o || pk_at(1) == CP_O)) {
+      adv(); adv()
+      while (pos < len && pk() >= CP_0 && pk() <= CP_7) adv()
+    } else {
+      while (pos < len && (is_digit(pk()) || pk() == CP_UNDERSCORE)) adv()
+      if (pos < len && pk() == CP_DOT) {
+        adv()
+        while (pos < len && (is_digit(pk()) || pk() == CP_UNDERSCORE)) adv()
+      }
+      if (pos < len && (pk() == CP_e || pk() == CP_E)) {
+        adv()
+        if (pos < len && (pk() == CP_PLUS || pk() == CP_MINUS)) adv()
+        while (pos < len && is_digit(pk())) adv()
+      }
+    }
+    raw = substr(start, pos)
+    push(tokens, {
+      kind: "number", at: start,
+      from_row: start_row, from_column: start_col,
+      to_row: row, to_column: col,
+      value: raw, number: number(raw)
+    })
+  }
+
+  var read_name = function() {
+    var start = pos
+    var start_row = row
+    var start_col = col
+    var name = ""
+    var kw = null
+    while (pos < len && is_ident_char(pk())) adv()
+    name = substr(start, pos)
+    kw = keywords[name]
+    if (kw != null) {
+      push(tokens, {
+        kind: kw, at: start,
+        from_row: start_row, from_column: start_col,
+        to_row: row, to_column: col
+      })
+    } else {
+      push(tokens, {
+        kind: "name", at: start,
+        from_row: start_row, from_column: start_col,
+        to_row: row, to_column: col,
+        value: name
+      })
+    }
+  }
+
+  var read_comment = function() {
+    var start = pos
+    var start_row = row
+    var start_col = col
+    var raw = ""
+    if (pk_at(1) == CP_SLASH) {
+      while (pos < len && pk() != CP_LF && pk() != CP_CR) adv()
+    } else {
+      adv(); adv() // skip /*
+      while (pos < len) {
+        if (pk() == CP_STAR && pk_at(1) == CP_SLASH) {
+          adv(); adv()
+          break
+        }
+        adv()
+      }
+    }
+    raw = substr(start, pos)
+    push(tokens, {
+      kind: "comment", at: start,
+      from_row: start_row, from_column: start_col,
+      to_row: row, to_column: col,
+      value: raw
+    })
+  }
+
+  var emit_op = function(kind, count) {
+    var start = pos
+    var start_row = row
+    var start_col = col
+    var i = 0
+    while (i < count) { adv(); i = i + 1 }
+    push(tokens, {
+      kind: kind, at: start,
+      from_row: start_row, from_column: start_col,
+      to_row: row, to_column: col
+    })
+  }
+
+  var emit_ident = function(count) {
+    var start = pos
+    var start_row = row
+    var start_col = col
+    var val = ""
+    var i = 0
+    while (i < count) { val = val + character(adv()); i = i + 1 }
+    push(tokens, {
+      kind: "name", at: start,
+      from_row: start_row, from_column: start_col,
+      to_row: row, to_column: col,
+      value: val
+    })
+  }
+
+  var tokenize_one = function() {
+    var c = pk()
+    var start = 0
+    var start_row = 0
+    var start_col = 0
+    var raw = ""
+    if (c == -1) return false
+
+    if (c == CP_LF) {
+      start = pos; start_row = row; start_col = col
+      adv()
+      push(tokens, { kind: "newline", at: start, from_row: start_row, from_column: start_col, to_row: row, to_column: col, value: "\n" })
+      return true
+    }
+    if (c == CP_CR) {
+      start = pos; start_row = row; start_col = col
+      adv()
+      if (pos < len && pk() == CP_LF) adv()
+      push(tokens, { kind: "newline", at: start, from_row: start_row, from_column: start_col, to_row: row, to_column: col, value: "\n" })
+      return true
+    }
+    if (c == CP_SPACE || c == CP_TAB) {
+      start = pos; start_row = row; start_col = col
+      while (pos < len && (pk() == CP_SPACE || pk() == CP_TAB)) adv()
+      raw = substr(start, pos)
+      push(tokens, { kind: "space", at: start, from_row: start_row, from_column: start_col, to_row: row, to_column: col, value: raw })
+      return true
+    }
+    if (c == CP_SQUOTE || c == CP_DQUOTE) { read_string(c); return true }
+    if (c == CP_BACKTICK) { read_template(); return true }
+    if (is_digit(c)) { read_number(); return true }
+    if (c == CP_DOT && is_digit(pk_at(1))) { read_number(); return true }
+    if (is_ident_start(c)) { read_name(); return true }
+    if (c == CP_SLASH) {
+      if (pk_at(1) == CP_SLASH || pk_at(1) == CP_STAR) { read_comment(); return true }
+      if (pk_at(1) == CP_EQ) { emit_op("/=", 2); return true }
+      if (pk_at(1) == CP_BANG) { emit_ident(2); return true }
+      emit_op("/", 1); return true
+    }
+    if (c == CP_STAR) {
+      if (pk_at(1) == CP_STAR) {
+        if (pk_at(2) == CP_BANG) { emit_ident(3); return true }
+        if (pk_at(2) == CP_EQ) { emit_op("**=", 3); return true }
+        emit_op("**", 2); return true
+      }
+      if (pk_at(1) == CP_EQ) { emit_op("*=", 2); return true }
+      if (pk_at(1) == CP_BANG) { emit_ident(2); return true }
+      emit_op("*", 1); return true
+    }
+    if (c == CP_PERCENT) {
+      if (pk_at(1) == CP_EQ) { emit_op("%=", 2); return true }
+      if (pk_at(1) == CP_BANG) { emit_ident(2); return true }
+      emit_op("%", 1); return true
+    }
+    if (c == CP_PLUS) {
+      if (pk_at(1) == CP_EQ) { emit_op("+=", 2); return true }
+      if (pk_at(1) == CP_PLUS) { emit_op("++", 2); return true }
+      if (pk_at(1) == CP_BANG) { emit_ident(2); return true }
+      emit_op("+", 1); return true
+    }
+    if (c == CP_MINUS) {
+      if (pk_at(1) == CP_EQ) { emit_op("-=", 2); return true }
+      if (pk_at(1) == CP_MINUS) { emit_op("--", 2); return true }
+      if (pk_at(1) == CP_BANG) { emit_ident(2); return true }
+      emit_op("-", 1); return true
+    }
+    if (c == CP_LT) {
+      if (pk_at(1) == CP_EQ && pk_at(2) == CP_BANG) { emit_ident(3); return true }
+      if (pk_at(1) == CP_EQ) { emit_op("<=", 2); return true }
+      if (pk_at(1) == CP_LT) {
+        if (pk_at(2) == CP_BANG) { emit_ident(3); return true }
+        if (pk_at(2) == CP_EQ) { emit_op("<<=", 3); return true }
+        emit_op("<<", 2); return true
+      }
+      if (pk_at(1) == CP_BANG) { emit_ident(2); return true }
+      emit_op("<", 1); return true
+    }
+    if (c == CP_GT) {
+      if (pk_at(1) == CP_EQ && pk_at(2) == CP_BANG) { emit_ident(3); return true }
+      if (pk_at(1) == CP_EQ) { emit_op(">=", 2); return true }
+      if (pk_at(1) == CP_GT) {
+        if (pk_at(2) == CP_GT) {
+          if (pk_at(3) == CP_BANG) { emit_ident(4); return true }
+          if (pk_at(3) == CP_EQ) { emit_op(">>>=", 4); return true }
+          emit_op(">>>", 3); return true
+        }
+        if (pk_at(2) == CP_BANG) { emit_ident(3); return true }
+        if (pk_at(2) == CP_EQ) { emit_op(">>=", 3); return true }
+        emit_op(">>", 2); return true
+      }
+      if (pk_at(1) == CP_BANG) { emit_ident(2); return true }
+      emit_op(">", 1); return true
+    }
+    if (c == CP_EQ) {
+      if (pk_at(1) == CP_EQ) {
+        if (pk_at(2) == CP_EQ) { emit_op("===", 3); return true }
+        emit_op("==", 2); return true
+      }
+      if (pk_at(1) == CP_GT) { emit_op("=>", 2); return true }
+      if (pk_at(1) == CP_BANG) { emit_ident(2); return true }
+      emit_op("=", 1); return true
+    }
+    if (c == CP_BANG) {
+      if (pk_at(1) == CP_EQ) {
+        if (pk_at(2) == CP_BANG) { emit_ident(3); return true }
+        if (pk_at(2) == CP_EQ) { emit_op("!==", 3); return true }
+        emit_op("!=", 2); return true
+      }
+      emit_op("!", 1); return true
+    }
+    if (c == CP_AMP) {
+      if (pk_at(1) == CP_AMP) {
+        if (pk_at(2) == CP_BANG) { emit_ident(3); return true }
+        if (pk_at(2) == CP_EQ) { emit_op("&&=", 3); return true }
+        emit_op("&&", 2); return true
+      }
+      if (pk_at(1) == CP_EQ) { emit_op("&=", 2); return true }
+      if (pk_at(1) == CP_BANG) { emit_ident(2); return true }
+      emit_op("&", 1); return true
+    }
+    if (c == CP_PIPE) {
+      if (pk_at(1) == CP_PIPE) {
+        if (pk_at(2) == CP_BANG) { emit_ident(3); return true }
+        if (pk_at(2) == CP_EQ) { emit_op("||=", 3); return true }
+        emit_op("||", 2); return true
+      }
+      if (pk_at(1) == CP_EQ) { emit_op("|=", 2); return true }
+      if (pk_at(1) == CP_BANG) { emit_ident(2); return true }
+      emit_op("|", 1); return true
+    }
+    if (c == CP_CARET) {
+      if (pk_at(1) == CP_EQ) { emit_op("^=", 2); return true }
+      if (pk_at(1) == CP_BANG) { emit_ident(2); return true }
+      emit_op("^", 1); return true
+    }
+    if (c == CP_LBRACKET) {
+      if (pk_at(1) == CP_RBRACKET && pk_at(2) == CP_BANG) { emit_ident(3); return true }
+      emit_op("[", 1); return true
+    }
+    if (c == CP_TILDE) {
+      if (pk_at(1) == CP_BANG) { emit_ident(2); return true }
+      emit_op("~", 1); return true
+    }
+    emit_op(character(c), 1)
+    return true
+  }
+
+  // Main loop
+  while (pos < len) {
+    tokenize_one()
+  }
+
+  // EOF token
+  push(tokens, { kind: "eof", at: pos, from_row: row, from_column: col, to_row: row, to_column: col })
+
+  return {filename: filename, tokens: tokens, cp: cp}
+}
+
+return tokenize

vm_test/label_break.txt

@@ -1 +0,0 @@
-var x = 0; outer: { x = 1; break outer; x = 2 }; x

vm_test/op_typeof.txt

@@ -1 +0,0 @@
-typeof 5

vm_test/op_void.txt

@@ -1 +0,0 @@
-void 0

vm_test/switch_basic.txt

@@ -1 +0,0 @@
-var r; switch(2) { case 1: r = 1; break; case 2: r = 2; break }; r

vm_test/syntax.txt

@@ -0,0 +1,336 @@
+/* comprehensive syntax test - exercises all AST paths */
+
+// --- variables and constants ---
+var a
+var b = 1
+var c = 2, d = 3
+def PI = 3.14159
+def MSG = "hello"
+
+// --- number literals ---
+var n_int = 42
+var n_float = 3.14
+var n_hex = 0xFF
+var n_octal = 0o77
+var n_binary = 0b1010
+var n_underscore = 1_000_000
+var n_exp = 1e10
+var n_exp_neg = 2.5e-3
+
+// --- string literals ---
+var s_single = 'single'
+var s_double = "double"
+var s_escape = "tab\there\nnewline\\slash\"quote"
+var s_unicode = "\u0041\u0042"
+
+// --- template literals ---
+var t_basic = `hello ${b} world`
+var t_nested = `a ${b + c} b ${d}`
+var t_expr = `result: ${b > 0 ? "yes" : "no"}`
+
+// --- array literals ---
+var arr_empty = []
+var arr_simple = [1, 2, 3]
+var arr_mixed = [1, "two", true, null]
+var arr_nested = [[1, 2], [3, [4, 5]]]
+
+// --- record/object literals ---
+var rec_empty = {}
+var rec_simple = {x: 1, y: 2}
+var rec_nested = {a: {b: {c: 3}}}
+var shorthand_var = 10
+var rec_shorthand = {shorthand_var}
+var rec_computed = {["key" + "1"]: 100}
+var rec_method = {
+  greet() { return "hi" }
+}
+var rec_mixed = {
+  name: "test",
+  value: 42,
+  nested: {inner: true},
+  items: [1, 2, 3]
+}
+
+// --- all binary arithmetic operators ---
+var arith_add = 1 + 2
+var arith_sub = 5 - 3
+var arith_mul = 4 * 5
+var arith_div = 10 / 3
+var arith_mod = 10 % 3
+var arith_pow = 2 ** 8
+
+// --- comparison operators ---
+var cmp_lt = 1 < 2
+var cmp_gt = 2 > 1
+var cmp_lte = 1 <= 1
+var cmp_gte = 2 >= 2
+var cmp_eq = 1 == 1
+var cmp_neq = 1 != 2
+
+// --- logical operators ---
+var log_and = true && false
+var log_or = false || true
+var log_not = !false
+
+// --- bitwise operators ---
+var bit_and = 0xFF & 0x0F
+var bit_or = 0xF0 | 0x0F
+var bit_xor = 0xFF ^ 0x0F
+var bit_not = ~0
+var bit_shl = 1 << 8
+var bit_shr = 256 >> 4
+var bit_shru = -1 >>> 0
+
+// --- nullish coalescing ---
+var nullish = null ?? "default"
+
+// --- ternary operator ---
+var tern = b > 0 ? "pos" : "neg"
+var tern_nested = b > 0 ? (b > 10 ? "big" : "small") : "neg"
+
+// --- comma operator ---
+var comma_result = (1, 2, 3)
+
+// --- unary operators ---
+var unary_pos = +b
+var unary_neg = -b
+
+// --- increment/decrement ---
+var inc_val = 0
+inc_val++
+inc_val--
+;++inc_val
+;--inc_val
+
+// --- all compound assignments ---
+var ca = 10
+ca += 5
+ca -= 3
+ca *= 2
+ca /= 4
+ca %= 3
+ca **= 2
+ca <<= 1
+ca >>= 1
+ca >>>= 0
+ca &= 0xFF
+ca |= 0x01
+ca ^= 0x10
+ca &&= true
+ca ||= false
+ca ??= 1
+
+// --- chained assignment ---
+var ch1, ch2
+ch1 = ch2 = 42
+
+// --- property access ---
+var obj = {a: {b: {c: 1}}, items: [10, 20, 30]}
+var dot_access = obj.a
+var dot_chain = obj.a.b.c
+var bracket_access = obj["a"]
+var bracket_dynamic = obj["it" + "ems"]
+var bracket_index = obj.items[1]
+
+// --- optional chaining ---
+var opt_obj = {x: {y: 1}}
+var opt_prop = opt_obj?.x
+var opt_deep = opt_obj?.x?.y
+var opt_null = null?.foo
+var opt_bracket = opt_obj?.["x"]
+var opt_call_obj = {f: function() { return 1 }}
+var opt_call = opt_call_obj?.f?.()
+
+// --- function declarations ---
+function add(a, b) {
+  return a + b
+}
+
+function no_params() {
+  return 42
+}
+
+function with_default(x = 10, y = 20) {
+  return x + y
+}
+
+// --- function expressions ---
+var func_expr = function(x) { return x * 2 }
+var func_named = function multiply(x, y) { return x * y }
+
+// --- arrow functions (all forms) ---
+var arrow_no_param = () => 0
+var arrow_one = x => x + 1
+var arrow_multi = (a, b) => a + b
+var arrow_block = (a, b) => { return a + b }
+var arrow_default = (x = 1, y = 2) => x + y
+var arrow_default_expr = (x = 1 + 2, y = arr_simple[0]) => x + y
+
+// --- closures ---
+function make_counter() {
+  var count = 0
+  return function() {
+    count += 1
+    return count
+  }
+}
+
+function outer_fn() {
+  var x = 10
+  function middle() {
+    var y = 20
+    function inner() {
+      return x + y
+    }
+    return inner()
+  }
+  return middle()
+}
+
+// --- this in methods ---
+var counter = {
+  val: 0,
+  inc() { this.val += 1 },
+  get() { return this.val }
+}
+
+// --- if/else/else-if ---
+var if_result
+if (b > 0) {
+  if_result = "positive"
+} else if (b == 0) {
+  if_result = "zero"
+} else {
+  if_result = "negative"
+}
+
+// --- while loop ---
+var w = 0
+while (w < 5) {
+  w += 1
+}
+
+// --- do-while loop ---
+var dw = 0
+do {
+  dw += 1
+} while (dw < 3)
+
+// --- for loop ---
+var f_sum = 0
+for (var i = 0; i < 10; i++) {
+  f_sum += i
+}
+
+// --- for loop with break ---
+var fb = 0
+for (var j = 0; j < 100; j++) {
+  if (j == 5) break
+  fb = j
+}
+
+// --- for loop with continue ---
+var fc = 0
+for (var k = 0; k < 10; k++) {
+  if (k % 2 == 0) continue
+  fc += k
+}
+
+// --- labeled break ---
+outer: for (var x = 0; x < 3; x++) {
+  for (var y = 0; y < 3; y++) {
+    if (y == 1) break outer
+  }
+}
+
+// --- labeled continue ---
+var lc = 0
+loop: for (var m = 0; m < 3; m++) {
+  for (var n = 0; n < 3; n++) {
+    if (n == 1) continue loop
+    lc += 1
+  }
+}
+
+// --- disrupt and disruption ---
+function disrupt_test() {
+  disrupt
+}
+
+function disruption_test() {
+  var x = 1
+} disruption {
+  var y = 2
+}
+
+function disrupt_with_disruption() {
+  disrupt
+} disruption {
+  var handled = true
+}
+
+// --- delete operator ---
+var del_obj = {a: 1, b: 2}
+delete del_obj.a
+
+// --- in operator ---
+var in_result = "b" in del_obj
+
+// --- go statement ---
+function async_task() { return 1 }
+function caller() { go async_task() }
+
+// --- IIFE ---
+var iife = (function() { return 99 })()
+
+// --- recursive function ---
+function factorial(n) {
+  if (n <= 1) return 1
+  return n * factorial(n - 1)
+}
+
+// --- mutually recursive functions ---
+function is_even(n) {
+  if (n == 0) return true
+  return is_odd(n - 1)
+}
+function is_odd(n) {
+  if (n == 0) return false
+  return is_even(n - 1)
+}
+
+// --- block scoping ---
+var block_val = 1
+{
+  var block_val = 2
+}
+
+// --- nested blocks ---
+{
+  var nb = 1
+  {
+    var nb2 = nb + 1
+    {
+      var nb3 = nb2 + 1
+    }
+  }
+}
+
+// --- empty statement ---
+;
+
+// --- parenthesized expression for precedence ---
+var prec = (1 + 2) * (3 + 4)
+
+// --- complex expressions ---
+var complex = arr_simple[0] + rec_simple.x * (b > 0 ? 2 : 1)
+
+// --- operator precedence chain ---
+var prec_chain = 1 + 2 * 3 ** 2 - 4 / 2 % 3
+
+// --- regex literals ---
+var re = /hello/
+var re_flags = /world/gi
+
+// --- line comment at end ---
+var end = 1 // done

vm_test/try_catch.txt

@@ -1 +0,0 @@
-try { 1 } catch(e) { 2 }

vm_test/try_catch_finally.txt

@@ -1 +0,0 @@
-var x = 1; try { throw 0 } catch(e) { x = 2 } finally { x = x + 1 }; x

vm_test/try_finally.txt

@@ -1 +0,0 @@
-var x = 1; try { x = 2 } finally { x = 3 }; x

vm_test/try_throw.txt

@@ -1 +0,0 @@
-try { throw "err" } catch(e) { e }

website/content/_index.md

@@ -0,0 +1,40 @@
+---
+title: "ƿit"
+---
+
+```javascript
+// hello.ce — a simple actor
+print("Hello, ƿit!")
+$stop()
+```
+
+```bash
+pit hello
+```
+
+## Why ƿit
+
+- **Actors, not threads** — isolated memory, message passing, no shared state. Concurrent programs that are safe by default.
+- **Everything is stoned** — `stone()` makes values permanently immutable. Messages between actors are frozen automatically. No defensive copying.
+- **Prototypes, not classes** — objects inherit directly from other objects. No class hierarchies, no `new`, no `this` confusion.
+- **C when you need it** — drop a `.c` file in your package and it becomes a native module. No FFI bindings, no build scripts.
+- **Small and predictable** — DEC64 numbers with no rounding errors. No `undefined`. Strict equality only. A runtime that fits in your head.
+
+<div class="home-art">
+  <img src="/images/wizard.png" alt="ƿit wizard">
+</div>
+
+<div class="home-links">
+  <a href="/start/">Get Started</a>
+  <a href="/manual/">Language Manual</a>
+</div>
+
+<div class="donate-section">
+
+## Support ƿit
+
+ƿit is free and open source. If you find it useful, consider supporting its development.
+
+Donation options coming soon.
+
+</div>

website/content/cli/_index.md

@@ -0,0 +1,175 @@
+---
+title: "Command Line Interface"
+description: "The pit tool"
+type: "standalone"
+---
+
+ƿit provides a command-line interface for managing packages, running scripts, and building applications.
+
+## Basic Usage
+
+```bash
+pit <command> [arguments]
+```
+
+## Commands
+
+### pit version
+
+Display the ƿit version.
+
+```bash
+pit version
+# 0.1.0
+```
+
+### pit install
+
+Install a package to the shop.
+
+```bash
+pit install gitea.pockle.world/john/prosperon
+pit install /Users/john/local/mypackage  # local path
+```
+
+### pit update
+
+Update packages from remote sources.
+
+```bash
+pit update              # update all packages
+pit update <package>    # update specific package
+```
+
+### pit remove
+
+Remove a package from the shop.
+
+```bash
+pit remove gitea.pockle.world/john/oldpackage
+```
+
+### pit list
+
+List installed packages.
+
+```bash
+pit list              # list all installed packages
+pit list <package>    # list dependencies of a package
+```
+
+### pit ls
+
+List modules and actors in a package.
+
+```bash
+pit ls              # list files in current project
+pit ls <package>    # list files in specified package
+```
+
+### pit build
+
+Build the current package.
+
+```bash
+pit build
+```
+
+### pit test
+
+Run tests.
+
+```bash
+pit test            # run tests in current package
+pit test all        # run all tests
+pit test <package>  # run tests in specific package
+```
+
+### pit link
+
+Manage local package links for development.
+
+```bash
+pit link add <canonical> <local_path>  # link a package
+pit link list                          # show all links
+pit link delete <canonical>            # remove a link
+pit link clear                         # remove all links
+```
+
+### pit fetch
+
+Fetch package sources without extracting.
+
+```bash
+pit fetch <package>
+```
+
+### pit upgrade
+
+Upgrade the ƿit installation itself.
+
+```bash
+pit upgrade
+```
+
+### pit clean
+
+Clean build artifacts.
+
+```bash
+pit clean
+```
+
+### pit help
+
+Display help information.
+
+```bash
+pit help
+pit help <command>
+```
+
+## Running Scripts
+
+Any `.ce` file in the ƿit core can be run as a command:
+
+```bash
+pit version    # runs version.ce
+pit build      # runs build.ce
+pit test       # runs test.ce
+```
+
+## Package Locators
+
+Packages are identified by locators:
+
+- **Remote**: `gitea.pockle.world/user/repo`
+- **Local**: `/absolute/path/to/package`
+
+```bash
+pit install gitea.pockle.world/john/prosperon
+pit install /Users/john/work/mylib
+```
+
+## Configuration
+
+ƿit stores its data in `~/.pit/`:
+
+```
+~/.pit/
+├── packages/       # installed packages
+├── lib/            # compiled dynamic libraries
+├── build/          # build cache
+├── cache/          # downloaded archives
+├── lock.toml       # installed package versions
+└── link.toml       # local development links
+```
+
+## Environment
+
+ƿit reads the `HOME` environment variable to locate the shop directory.
+
+## Exit Codes
+
+- `0` — Success
+- Non-zero — Error (check output for details)

website/content/contributing/_index.md

@@ -0,0 +1,60 @@
+---
+title: "Contributing"
+description: "How to contribute to ƿit"
+type: "standalone"
+---
+
+ƿit is developed openly. Contributions of all kinds are welcome.
+
+## Report Bugs
+
+Found a problem? Open an issue on the [ƿit issue tracker](https://gitea.pockle.world/john/cell/issues). Include:
+
+- What you expected to happen
+- What actually happened
+- A minimal reproduction (a short `.ce` or `.cm` file)
+- Your platform and ƿit version (`pit version`)
+
+## Submit Packages
+
+Share your ƿit packages by hosting them on a Gitea instance. Any package with a valid `pit.toml` can be installed by others:
+
+```bash
+pit install gitea.example.com/you/your-package
+```
+
+See [Packages](/manual/#packages) for how to structure and publish packages.
+
+## Contribute to the Runtime
+
+The ƿit runtime is written in C. To build from source:
+
+```bash
+git clone https://gitea.pockle.world/john/cell
+cd cell
+make bootstrap
+```
+
+### Code Style
+
+- C code uses 2-space indentation
+- Functions and variables are `static` unless exported
+- No headers between files in the same package
+- Use `JS_NULL` / `JS_IsNull` — there is no `undefined`
+- Objects over classes; limit prototype usage
+
+### Submitting Patches
+
+1. Fork the repository on Gitea
+2. Create a branch for your change
+3. Keep commits focused — one logical change per commit
+4. Test your changes with `pit test all`
+5. Open a pull request with a clear description
+
+## Improve Documentation
+
+Documentation lives in the `docs/` directory as Markdown files. Fixes for typos, unclear explanations, or missing examples are always appreciated.
+
+## Code of Conduct
+
+Be respectful. Communicate clearly. Assume good faith. Technical disagreements are fine; personal attacks are not.