@kudzujs/core 0.8.43 → 0.8.45

package/MIGRATION_ROADMAP.md

@@ -2,7 +2,7 @@
 
 This document is the source of truth for Kudzu's product direction, architecture invariants, and future development order. Read it before extending React-shaped syntax or browser capabilities.
 
-The executable post-`0.8.43` compiler and large-application sequence is maintained in [`docs/next-architecture/large-application-ai-native-roadmap.md`](./docs/next-architecture/large-application-ai-native-roadmap.md). Follow its PR dependencies for implementation work; this document remains authoritative when selecting or accepting a migration capability.
+The executable post-`0.8.45` compiler and large-application sequence is maintained in [`docs/next-architecture/large-application-ai-native-roadmap.md`](./docs/next-architecture/large-application-ai-native-roadmap.md). Follow its PR dependencies for implementation work; this document remains authoritative when selecting or accepting a migration capability.
 
 [`GOAL_A.md`](./GOAL_A.md) and [`GOAL_B.md`](./GOAL_B.md) are completed capability-validation records. Their commerce and realtime dashboard fixtures prove general lifecycle, navigation, async-workflow, and Worker capabilities; they are not separate product verticals or future priority lists.
 

package/PERFORMANCE.md

@@ -1,8 +1,20 @@
 # Performance Records
 
-Reproducibility classes: `npm run benchmark`, `npm run benchmark:keyed`, `npm run benchmark:native`, and `npm run benchmark:module-cache` are maintained in this repository; `npm run benchmark:commerce` is a maintained paired runner over the public external storefront; older excluded-workspace sections are historical provenance only and are not current framework rankings.
+Reproducibility classes: `npm run benchmark`, `npm run benchmark:keyed`, `npm run benchmark:native`, `npm run benchmark:module-cache`, and `npm run benchmark:source-scale` are maintained in this repository; `npm run benchmark:commerce` is a maintained paired runner over the public external storefront; older excluded-workspace sections are historical provenance only and are not current framework rankings.
 
-## Current 0.8.43 Release Snapshot
+`npm run benchmark:source-scale` generates its fixture outside the repository so 50,000 lines of synthetic source are not tracked. The default topology is 50 pages plus 450 route-owned imported modules. Generation is excluded from timing; fresh-process samples separately report source reads, reachable-graph discovery, source compilation, clean production build, compiler-result and deploy digests, output files/bytes, cache counters, and peak RSS. `ROUTES`, `MODULES_PER_ROUTE`, `FILLER_LINES`, `WARMUPS`, and `RUNS` may reduce or expand the fixture without changing the default acceptance floor. `TARGET_ROOT` measures another checkout; `BASELINE_ROOT` alternates that checkout with the current tree and requires identical deploy output.
+
+The maintained 2026-08-13 comparison used Node 24.14.0 and an Intel Core i5-9500 Linux x64 host, one warm-up, and seven alternating fresh-process samples against clean `v0.8.44`. A narrow fast path skips Kudzu semantic transformation for 450 plain `.ts` modules whose runtime edges are exclusively resolvable relative TypeScript imports or exports; all other modules retain the existing transformer. Compile median fell from 2,323.9 ms to 1,413.2 ms (39.2%) and clean-build median from 3,325.3 ms to 2,382.4 ms (28.4%); every paired sample improved. Compile peak-RSS median fell from 571.2 MiB to 552.6 MiB, while build peak RSS was 570.9 MiB versus 568.8 MiB. Compiler scratch fell from 7,328,390 to 1,971,061 bytes. Both targets emitted the same 50 static HTML files, 10,980 bytes, and deploy SHA-256 `e107d78a7f55bc8a1af0ea6e53efeffa19b3d44d21c892484d103fa346e7ba7b`. This is a source-scale compiler comparison, not a cross-framework result.
+
+## Current 0.8.45 Optimization Snapshot
+
+Kudzu 0.8.45 adds the plain TypeScript fast path and maintained paired source-scale runner described above. It changes compiler scratch only: the measured deploy graph is byte-identical to `v0.8.44`, and no browser runtime or public API changed.
+
+## Maintained 0.8.44 Release Snapshot
+
+Kudzu 0.8.44 changes compiler-only naming for action-private Context state and setters. The maintained Context browser fixture retains the same emitted concrete state operations and CRUD behavior while a same-named consumer local remains ordinary static content. No browser runtime module or public API changed, and no new performance claim is made.
+
+## Maintained 0.8.43 Release Snapshot
 
 Kudzu 0.8.43 extends compiler-only callback/ref specialization from two to three direct component boundaries. The maintained browser fixture retains the same parent state operations, child state/effect/ref ownership, conditional cleanup, and static zero-JavaScript sibling while the added forwarding component is absent from emitted JavaScript. No browser runtime module or public API changed, and no new performance ranking is claimed.
 

package/README.md

@@ -14,7 +14,7 @@ Kudzu compiles ordinary React-shaped TypeScript and TSX into complete static HTM
 
 > Experimental `0.8.x`: the compiler API and supported TSX surface may change.
 
-**Latest release: 0.8.43 - Three-boundary callback ownership.** Direct setters, simple state callbacks, and object refs can now cross a child plus two direct forwarding components while Kudzu preserves parent signals, child hooks, and DOM ownership without retaining component functions. Read the [release notes](./RELEASES.md#0843---three-boundary-callback-ownership), open the [release page](https://github.com/kudzujs/kudzu/releases/tag/v0.8.43), or follow the [architecture packet](./docs/next-architecture/README.md).
+**Latest release: 0.8.45 - Plain TypeScript fast path.** Proven relative-only `.ts` modules skip Kudzu's TSX semantic transformer while uncertain modules retain the existing path. The maintained 500-module fixture reduced compile median 39.2% and clean-build median 28.4% with identical deploy output. Read the [release notes](./RELEASES.md#0845---plain-typescript-fast-path), open the [release page](https://github.com/kudzujs/kudzu/releases/tag/v0.8.45), or follow the [architecture packet](./docs/next-architecture/README.md).
 
 - [Documentation](https://kudzujs.cloud/docs)
 - [Installation guide](https://kudzujs.cloud/docs#install)

package/RELEASES.md

@@ -1,5 +1,65 @@
 # Kudzu Releases
 
+## 0.8.45 - Plain TypeScript fast path
+
+Kudzu 0.8.45 reduces large-project compiler work by sending proven plain `.ts` modules through TypeScript transpilation and relative ESM path rewriting without running Kudzu's TSX semantic transformer.
+
+### Changed in 0.8.45
+
+- The fast path is limited to `.ts` files whose runtime imports and exports are resolvable relative TypeScript modules.
+- TSX, package imports, static assets, unresolved edges, and all uncertain module shapes retain the existing Kudzu transformer.
+- `npm run benchmark:source-scale` deterministically generates 500 reachable modules and supports alternating baseline/candidate measurements with deploy-output equivalence checks.
+- Compiler counters report the number of plain modules selected.
+
+### Output and performance
+
+- Seven alternating samples against clean `v0.8.44` reduced compile median from 2,323.9 ms to 1,413.2 ms (39.2%) and clean-build median from 3,325.3 ms to 2,382.4 ms (28.4%) on the maintained source-scale fixture.
+- Compiler scratch decreased from 7,328,390 to 1,971,061 bytes and compile peak-RSS median decreased from 571.2 MiB to 552.6 MiB.
+- Both targets emitted the same 50 HTML files, 10,980 deploy bytes, and SHA-256 digest. No browser runtime or public API changed.
+
+### Validation
+
+- `npm run check`, `npm test`, and `npm run test:package` pass with all 209 tests and 160 generated pages.
+- A reduced deterministic source-scale run remains in the standard suite.
+- The maintained benchmark alternates target order and rejects any deploy digest, file-count, page-count, or byte-count difference.
+
+### Upgrade
+
+```bash
+npm install @kudzujs/core@^0.8.45
+```
+
+## 0.8.44 - Collision-free Context actions
+
+Kudzu 0.8.44 removes a compiler-only naming restriction from relative Context-hook actions: Provider state and setter fields needed privately by an action may now share names with ordinary consumer locals.
+
+### Changed in 0.8.44
+
+- Action-required Provider state/setter fields that are not publicly selected receive compiler-owned collision-free aliases in the consumer destructuring.
+- The specialized action AST and its state map use the same aliases before existing HandlerIR lowering.
+- Alias selection checks component parameters and top-level declarations and increments the compiler-owned suffix when necessary.
+- Consumer source keeps its chosen local names and public hook destructuring; no application rename or extra Context field selection is required.
+- Provider state identity, direct state operations, and existing Context action diagnostics remain unchanged.
+
+### Output and performance
+
+- Context action functions, Context objects, and Provider trees remain absent from browser output.
+- Static siblings remain complete zero-JavaScript documents.
+- No browser runtime module or public API changed. This release makes no new performance claim.
+
+### Validation
+
+- `npm run check`, `npm test`, and `npm run test:package` pass with all 208 tests and 158 generated pages.
+- The maintained Notes Context fixture now declares a consumer-local `setNotes` while create, rename, delete, select, list identity, and active-note updates continue to pass in Chrome.
+- A focused counter fixture previously rejected for consumer `setCount` collision now builds successfully.
+- Private Provider captures, hidden state pairs, and indirect action references remain source-diagnosed.
+
+### Upgrade
+
+```bash
+npm install @kudzujs/core@^0.8.44
+```
+
 ## 0.8.43 - Three-boundary callback ownership
 
 Kudzu 0.8.43 extends ordinary React-shaped callback and ref composition through one more proven direct forwarding component without adding a callback registry, browser component tree, or runtime module.

package/docs/next-architecture/README.md

@@ -11,7 +11,7 @@ The top-level [`GOAL_A.md`](../../GOAL_A.md) and [`GOAL_B.md`](../../GOAL_B.md)
 | C: state/resource model | Research only | Reduced fixtures expose a limitation |
 | D: routing compatibility | Current behavior preserved | Revisit only with migration evidence and invariant review |
 
-The active post-`0.8.43` implementation sequence is [`large-application-ai-native-roadmap.md`](./large-application-ai-native-roadmap.md). P0.12 validation, property-level object-state dependencies, direct three-boundary callback/ref ownership, and the measured route-entry output optimization are complete; broader prop, callback, ref, and Context dataflow remains migration-backed work. The plan orders compiler semantic generalization, large-application foundations, compatibility boundaries, AI tooling, and production validation without changing the invariants below.
+The active post-`0.8.45` implementation sequence is [`large-application-ai-native-roadmap.md`](./large-application-ai-native-roadmap.md). P0.12 validation, property-level object-state dependencies, direct three-boundary callback/ref ownership, collision-free Context action-private state, measured route-entry output optimization, and the plain TypeScript source-scale fast path are complete; broader prop, callback, ref, and Context dataflow remains migration-backed work. The plan orders compiler semantic generalization, large-application foundations, compatibility boundaries, AI tooling, and production validation without changing the invariants below.
 
 ## Required Invariants
 

package/docs/next-architecture/compiler-current-architecture.md

@@ -1,6 +1,6 @@
 # Current Compiler Architecture
 
-This maps the current `0.8.43` architecture, built on the completed `0.8.23` Goal A compiler foundation. File and function names are the stable references; line numbers are intentionally omitted because later work may still move code.
+This maps the current `0.8.45` architecture, built on the completed `0.8.23` Goal A compiler foundation. File and function names are the stable references; line numbers are intentionally omitted because later work may still move code.
 
 ## Responsibility Map
 
@@ -17,7 +17,7 @@ This maps the current `0.8.43` architecture, built on the completed `0.8.23` Goa
 | Source-local binding index | [`framework/compiler/analysis/binding-index.mjs`](../../framework/compiler/analysis/binding-index.mjs) | After normalization, assigns deterministic lexical slots and classifies local, parameter, import, capture, global, and unresolved references. Native handler, effect, binding, list evaluator, optimized-command, and effect-resource consumers use complete index-owned AST; synthesized expressions retain the existing fallback. |
 | Pure collection language | [`framework/compiler/collection-analysis.mjs`](../../framework/compiler/collection-analysis.mjs) | Analyzes collection roots/selectors and serializes the allowed pure expression language used by lists and derived dependencies. |
 | Main semantic analysis | [`framework/compiler/source-compiler.mjs`](../../framework/compiler/source-compiler.mjs), `createKudzuTransformer()` | Produces transformed source plus explicit component, handler, binding, derived, keyed, and effect ownership results. |
-| Component ownership analysis | [`framework/compiler/analysis/component-analysis.mjs`](../../framework/compiler/analysis/component-analysis.mjs) | Produces ComponentAnalysis v2 with ordered JSON-safe owner and specialization slots for state, setters, props, refs, IDs, direct SignalIR links, structural OwnerRefs, source-local SiteIds, and source provenance. One additional proven callback/ref forwarding component specializes into the same parent signal and intrinsic ownership; AST identity remains private to its source-local session. |
+| Component ownership analysis | [`framework/compiler/analysis/component-analysis.mjs`](../../framework/compiler/analysis/component-analysis.mjs) | Produces ComponentAnalysis v2 with ordered JSON-safe owner and specialization slots for state, setters, props, refs, IDs, direct SignalIR links, structural OwnerRefs, source-local SiteIds, and source provenance. Three direct callback/ref component boundaries specialize into the same parent signal and intrinsic ownership; Context action-private state fields receive collision-free consumer-local aliases; AST identity remains private to its source-local session. |
 | Per-source descriptor registration | [`framework/compiler/descriptor-session.mjs`](../../framework/compiler/descriptor-session.mjs), `createSemanticArtifact()`, `createDescriptorSession()` | Keeps AST descriptors private during analysis, then finalizes ModuleIR v2 with deterministic SymbolRef, SignalIR, HandlerIR, BindingIR, DerivedIR, EffectIR, KeyedBlockIR, and ImportIR slots. One fail-closed boundary validates every source-local and component ownership edge before build-module generation. |
 | Route artifact graph | [`framework/compiler/route-build-record.mjs`](../../framework/compiler/route-build-record.mjs), `createRouteBuildRecord()`, `planRouteArtifacts()` | Validates each rendered route's RouteIR, capabilities, entry paths, styles, and exact handler/effect references. Handler modules, Workers, package-client modules, and bundle entry/chunk closure derive from structural edges without serialized output searches or formatted composite keys. |
 | Route contract validation | [`framework/compiler/route-ir.mjs`](../../framework/compiler/route-ir.mjs), `assertRouteIR()` | Fails before artifact selection for invalid state/parameter identity, commands, native/effect captures and dependencies, reactive descriptors, conditions, keyed-list identity/ownership, marker fields, or JSON safety. Immutable in-memory contracts validate once by identity. |

package/docs/next-architecture/large-application-ai-native-roadmap.md

@@ -141,6 +141,7 @@ This is an incremental evolution of the current repository:
 - [x] P0.10 ModuleIR Reference Unification is complete in `0.8.37`. ModuleIR and ComponentAnalysis v2 assign deterministic slots to symbols, signals, handlers, bindings, derived values, effects, keyed blocks, imports, owners, specializations, states, refs, and IDs. State, capture, import, effect, collection, parent/child, specialization, and row edges use structural slots or ModuleSymbol records while readable names remain codegen/debug metadata. A fail-closed pre-codegen validator rejects malformed slots, unsupported versions, duplicate exports, broken reciprocity, and ownership cycles; focused JSON round-trip checks and all 198 tests pass without changing browser runtime behavior.
 - [x] P0.11 RouteBuildRecord and Artifact Graph is complete in `0.8.38`. Each rendered route records RouteIR, capability facts, route-entry paths, styles, and exact handler/effect references. Build orchestration derives Handler ESM, Worker, package-client, and chunk closure from those edges; serialized HTML/plan `includes()` searches, formatted effect keys, and parallel route-fact/entry arrays are removed. Focused malformed-reference and JSON round-trip checks plus the standard suite preserve the exact 173-file deploy digest and bytes.
 - [x] P0.12 Deep RouteIR and CapabilityIR Validation is complete in `0.8.39`. RouteIR v1 validates state/parameter identity, commands, captures, effects, bindings, conditions, keyed-list ownership, and JSON safety before artifact selection. RouteBuildRecord checks concrete capability reciprocity, and CapabilityIR validates standalone implications plus exact projection from route records before codegen. Identity caching avoids repeated validation of immutable contracts; all 203 tests and byte-identical deploy checks pass without runtime or accepted-source changes.
+- [x] The P1 source-scale benchmark foundation is complete after `0.8.44`. `npm run benchmark:source-scale` generates 500 reachable modules, 50,550 TS/TSX lines, and 50 routes outside the repository, then measures source read, graph discovery, compile, clean build, deterministic output, cache counters, and peak RSS in fresh processes. Its paired mode alternates another checkout with the current tree and rejects deploy-output differences. The first measured specialization keeps Kudzu transformation for TSX, package imports, assets, and unresolved edges while plain relative-only `.ts` modules receive path rewriting alone; seven alternating samples against `v0.8.44` reduced compile median 39.2% and clean-build median 28.4% with identical deploy output. A reduced deterministic run remains in the standard suite.
 
 ### P0: Semantic Correctness And Compiler Foundation
 
@@ -175,7 +176,7 @@ P0 creates the semantic base required by every later large-application capabilit
 | Route capability and chunk closure | Every route reports exact runtime, handler, package, Worker, and shared chunk edges |
 | Route/layout CSS closure | Unrelated feature CSS is absent from a route unless configured global |
 | Incremental development build | A source change recompiles and rerenders only affected modules/routes while preserving full reload correctness |
-| Source-scale benchmark | At least 500 reachable modules, 50,000 TS/TSX lines, 50 routes, phase timings, and peak RSS |
+| Source-scale benchmark | `npm run benchmark:source-scale` deterministically generates at least 500 reachable modules, 50,000 TS/TSX lines, and 50 routes, then reports source-read, graph, compile, clean-build, output, and peak-RSS measurements from fresh processes |
 
 ### P2: Compatibility And Migration
 
@@ -378,7 +379,7 @@ function increment(value) { setCount(value + 1) }; increment(count)
 After the relevant P0 foundations, investigate capabilities in this order:
 
 1. Property-level derived dependencies over ordinary object state. **Completed in `0.8.40`:** direct property paths and top-level immutable primitive locals over object state reuse tagged DerivedIR, subscribe to the source signal, and compare selected values with `Object.is`; whole-object and dynamic dependencies remain rejected.
-2. Multi-boundary component/prop/callback/ref/context dataflow. **Three-boundary callback/ref ownership completed in `0.8.43`:** two additional direct forwarding components preserve the parent SignalIR and final intrinsic ownership. Each forwarding component must destructure and forward the callback directly once; a fourth boundary remains fail-closed. Broader prop, callback, ref, and Context graphs remain migration-led work.
+2. Multi-boundary component/prop/callback/ref/context dataflow. **Three-boundary callback/ref ownership completed in `0.8.43`; collision-free Context action-private state completed in `0.8.44`:** forwarding preserves parent SignalIR and Context action lowering uses compiler-owned aliases when consumer locals reuse Provider state/setter names. A fourth callback boundary remains fail-closed. Broader prop, callback, ref, and Context graphs remain migration-led work.
 3. Package-neutral shared state/actions and migration of current Zustand internals.
 4. Browser-only package imports in owned effect/resource modules.
 5. ResourceIR from at least two independent WebSocket/SSE/SDK fixtures with the same semantics.
@@ -480,4 +481,4 @@ The first comparison is Kudzu versus React + Vite using the same agent, model, t
 
 ## Immediate Decision
 
-PR 1 through PR 12, the `0.8.40` property-dependency slice, the `0.8.41` and `0.8.43` direct multi-boundary callback/ref slices, and the `0.8.42` measured route-output optimization are complete. Continue broader **component/prop/callback/ref/context dataflow** only from the next real migration blocker. Do not skip directly to a store, resource, router, virtualization, or ecosystem package feature.
+PR 1 through PR 12, the `0.8.40` property-dependency slice, the `0.8.41` and `0.8.43` direct multi-boundary callback/ref slices, the `0.8.44` Context alias slice, and the `0.8.42` measured route-output optimization are complete. Continue broader **component/prop/callback/ref/context dataflow** only from the next real migration blocker. Do not skip directly to a store, resource, router, virtualization, or ecosystem package feature.

package/docs/next-architecture/versioning.md

@@ -1,6 +1,6 @@
 # Planned Version Sequence
 
-This is an execution sequence, not release history. `0.8.16` through `0.8.43` are completed scopes represented by package/release records.
+This is an execution sequence, not release history. `0.8.16` through `0.8.45` are completed scopes represented by package/release records.
 
 Keep each patch behavior-preserving and independently reviewable. If a boundary proves inseparable, revise this plan before combining releases; do not silently broaden a patch.
 
@@ -34,6 +34,8 @@ Keep each patch behavior-preserving and independently reviewable. If a boundary
 | `0.8.41` | Preserve callback, setter, and ref ownership through one additional direct presentation-component boundary. | Parent signals, child hooks, IDs, effects, refs, conditional cleanup, static zero-JavaScript output, and byte-identical deploy graphs pass; a third callback boundary fails closed. |
 | `0.8.42` | Share byte-identical generated route entries and accelerate one proven read-only query form carry shape. | A static-path fixture emits one native/effect file for three routes; hidden GET carry initializes before deferred parameter ESM; static siblings remain zero-JavaScript; the losing catalog navigation experiment is excluded. |
 | `0.8.43` | Preserve callback, setter, and ref ownership through a third direct component boundary. | Parent SignalIR, child state/effects/IDs, object refs, conditional cleanup, fresh remount, and static zero-JavaScript output pass in Chrome; a fourth callback boundary fails closed. |
+| `0.8.44` | Remove consumer-local name collisions from specialized Context actions. | Action-private Provider state/setter fields receive deterministic collision-free aliases; CRUD browser behavior, direct state operations, diagnostics, and zero Context runtime output pass. |
+| `0.8.45` | Add a source-scale benchmark and skip Kudzu semantic transformation for proven plain TypeScript modules. | Seven alternating samples improve compile and clean-build medians with identical deploy output; uncertain modules retain the existing transformer. |
 
 ## Sequence Rules
 

package/framework/README.md

@@ -2,7 +2,7 @@
 
 Kudzu specializes ordinary common React-shaped TSX so migrations need minimal source restructuring. Declarative components, collection pipelines, conditions, hooks, and handlers should be lowered at build time rather than replaced with application-owned imperative DOM code. This principle applies across migrations and is not Stay-specific; it does not imply a React package, VDOM, hydration, or ecosystem runtime.
 
-Migration source may retain conventional `react` imports for supported named or aliased hooks, direct members such as `React.useState`, same-file `memo`, inline `useCallback`, direct-state expression or analyzable collection-pipeline `useMemo`, direct intrinsic `forwardRef`, top-level `const` identifiers initialized by `useId()`, and default, namespace, or named `Fragment`. A named or default zero-argument custom hook imported from a relative TypeScript module may return direct shorthand state/setter pairs and callbacks that capture those states, including a callback that resets multiple states to direct primitive literals; callers use one top-level `const` object destructuring without aliases, defaults, or rest. Kudzu's JSX declarations accept ReactNode-shaped component returns and contextually type common intrinsic DOM events, so strict React component props do not need migration-only `unknown` or explicit event annotations. `forwardRef()` accepts one inline synchronous `(props, ref)` function and requires the object ref exactly once on its direct intrinsic root; the compiler removes `ref` from props/rest and erases the wrapper. `useId()` becomes a deterministic build-time HTML ID and emits no browser capability; keyed rows reject it because cloned row templates cannot safely duplicate HTML IDs. Collection memos may start from local array state or a named relative import of an exported JSON-safe `const` array, including type-only `as const` and `satisfies` wrappers, and may read direct local state declared in their dependency array. `compiler/react-migration-pass.mjs` canonicalizes those forms and rewrites module references to `@kudzujs/core` before build-time evaluation. Memo wrappers are erased or inlined into existing bindings and keyed-list selectors because no browser component rerender or memo cache exists. Static routes remain JavaScript-free and emitted modules are checked for surviving React imports.
+Migration source may retain conventional `react` imports for supported named or aliased hooks, direct members such as `React.useState`, same-file `memo`, inline `useCallback`, direct-state expression or analyzable collection-pipeline `useMemo`, direct intrinsic `forwardRef`, top-level `const` identifiers initialized by `useId()`, and default, namespace, or named `Fragment`. A named or default zero-argument custom hook imported from a relative TypeScript module may return direct shorthand state/setter pairs and callbacks that capture those states, including a callback that resets multiple states to direct primitive literals; callers use one top-level `const` object destructuring without aliases, defaults, or rest. Relative Context-hook actions may require exposed Provider state/setter fields without forcing those hidden fields into public consumer destructuring; compiler-owned collision-free aliases preserve same-named consumer locals. Kudzu's JSX declarations accept ReactNode-shaped component returns and contextually type common intrinsic DOM events, so strict React component props do not need migration-only `unknown` or explicit event annotations. `forwardRef()` accepts one inline synchronous `(props, ref)` function and requires the object ref exactly once on its direct intrinsic root; the compiler removes `ref` from props/rest and erases the wrapper. `useId()` becomes a deterministic build-time HTML ID and emits no browser capability; keyed rows reject it because cloned row templates cannot safely duplicate HTML IDs. Collection memos may start from local array state or a named relative import of an exported JSON-safe `const` array, including type-only `as const` and `satisfies` wrappers, and may read direct local state declared in their dependency array. `compiler/react-migration-pass.mjs` canonicalizes those forms and rewrites module references to `@kudzujs/core` before build-time evaluation. Memo wrappers are erased or inlined into existing bindings and keyed-list selectors because no browser component rerender or memo cache exists. Static routes remain JavaScript-free and emitted modules are checked for surviving React imports.
 
 Compilation begins from page entries and follows relative runtime imports, re-exports, and validated Worker references; unreachable TypeScript migration files are not transformed. Direct maps over imported immutable JSON-safe arrays fold to literals for zero-JavaScript static rows. Synchronous relative calculation functions may return objects whose direct static fields feed reactive JSX bindings; build rendering uses current signal values and route-specific binding ESM reevaluates the same helper after state commits. One direct array field may instead feed a keyed intrinsic map: its evaluator refreshes a compiler-owned array anchor before the existing list reconciler runs, preserving keyed DOM and SVG identity without a calculation runtime. That field must remain a JSON-safe array after every source-state commit. Package imports have a separate narrow boundary: direct references inside intrinsic JSX event callbacks are erased from build modules and bundled into route handler ESM, while render-time, effect, helper-indirect, and mixed package use fails.
 

package/framework/compiler/project-session.mjs

@@ -11,6 +11,7 @@ export function createProjectSession(projectRoot = process.cwd(), { counters, so
   const graph = createSourceGraph(root)
   const modules = createModuleCache(root, sourceIndex, graph, counters)
   return {
+    counters,
     root,
     sourceDirectory,
     pagesDirectory: join(sourceDirectory, "pages"),

package/framework/compiler/source-compiler.mjs

@@ -22,7 +22,7 @@ import { createProjectSession } from "./project-session.mjs"
 import { createZustandPass } from "./zustand-pass.mjs"
 
 export function createSourceCompiler(project) {
-const { root, sourceDirectory, pagesDirectory, workDirectory, workerCompiler, modules } = project
+const { root, sourceDirectory, pagesDirectory, workDirectory, workerCompiler, modules, counters } = project
 const { ordinaryRuntimeDependencies, resolveSourceImport, runtimeModuleReference } = project.graph
 const parseSourceFile = (file, source) => modules.read(file, source).sourceFile
 const staticAssetExtensions = new Set([".avif", ".gif", ".ico", ".jpeg", ".jpg", ".otf", ".png", ".svg", ".ttf", ".webp", ".woff", ".woff2"])
@@ -32,6 +32,8 @@ function compileSource(file, sourceFiles, sourceIndex, staticFiles, cssModules,
   const source = sourceIndex.get(file)
   const semantic = createSemanticArtifact(relative(root, file).replaceAll(sep, "/"))
   const handlerPath = `handlers/${relative(sourceDirectory, file).replaceAll(sep, "/").replace(/\.(?:ts|tsx)$/, ".js")}`
+  const plain = plainTypeScriptModule(file, source, sourceFiles)
+  if (plain && counters) counters.plainModules = (counters.plainModules ?? 0) + 1
   const result = ts.transpileModule(source, {
     fileName: file,
     compilerOptions: {
@@ -40,7 +42,7 @@ function compileSource(file, sourceFiles, sourceIndex, staticFiles, cssModules,
       jsx: ts.JsxEmit.ReactJSX,
       jsxImportSource: "@kudzujs/core"
     },
-    transformers: { before: [createKudzuTransformer({ semantic, handlerUrl: assetPath(base, `assets/${handlerPath}`), file, sourceFiles, sourceIndex, staticFiles, importedAssets, cssModules, base })] },
+    transformers: { before: [plain ? createPlainModuleTransformer(file, sourceFiles) : createKudzuTransformer({ semantic, handlerUrl: assetPath(base, `assets/${handlerPath}`), file, sourceFiles, sourceIndex, staticFiles, importedAssets, cssModules, base })] },
     reportDiagnostics: true
   })
 
@@ -71,6 +73,27 @@ function compileSource(file, sourceFiles, sourceIndex, staticFiles, cssModules,
   return sourceResult
 }
 
+function plainTypeScriptModule(file, source, sourceFiles) {
+  if (!file.endsWith(".ts")) return false
+  const sourceFile = parseSourceFile(file, source)
+  for (const statement of sourceFile.statements) {
+    if ((!ts.isImportDeclaration(statement) && !ts.isExportDeclaration(statement)) || !runtimeModuleReference(statement)) continue
+    if (!statement.moduleSpecifier || !ts.isStringLiteral(statement.moduleSpecifier) || !statement.moduleSpecifier.text.startsWith(".") || isStaticImport(statement.moduleSpecifier.text)) return false
+    try { resolveSourceImport(file, statement.moduleSpecifier.text, sourceFiles) } catch { return false }
+  }
+  return true
+}
+
+function createPlainModuleTransformer(file, sourceFiles) {
+  return context => sourceFile => context.factory.updateSourceFile(sourceFile, sourceFile.statements.map(statement => {
+    if ((!ts.isImportDeclaration(statement) && !ts.isExportDeclaration(statement)) || !runtimeModuleReference(statement) || !statement.moduleSpecifier || !ts.isStringLiteral(statement.moduleSpecifier) || !statement.moduleSpecifier.text.startsWith(".")) return statement
+    const target = resolveSourceImport(file, statement.moduleSpecifier.text, sourceFiles)
+    const specifier = context.factory.createStringLiteral(relativeModulePath(compiledPath(file), compiledPath(target)))
+    if (ts.isImportDeclaration(statement)) return context.factory.updateImportDeclaration(statement, statement.modifiers, statement.importClause, specifier, statement.attributes)
+    return context.factory.updateExportDeclaration(statement, statement.modifiers, statement.isTypeOnly, statement.exportClause, specifier, statement.attributes)
+  }))
+}
+
 function normalizeModulePaths(moduleIR) {
   const normalize = target => isAbsolute(target) ? relative(root, target).replaceAll(sep, "/") : target
   moduleIR.imports = moduleIR.imports.map(entry => entry.package ? entry : { ...entry, target: normalize(entry.target) })
@@ -548,22 +571,33 @@ function createKudzuTransformer({ semantic, handlerUrl, file, sourceFiles, sourc
               if (callback) for (const state of referencedStateNames(callback.body, hook.states, callback)) requiredContextStates.add(state)
             }
           }
+          const contextSubstitutions = new Map()
           for (const [setter, state] of hook.states) {
             if (hook.context) {
               if (names.has(setter) && !names.has(state)) throw sourceNodeError(node.name, sourceFile, `Relative Context setter ${JSON.stringify(setter)} requires state ${JSON.stringify(state)} to be destructured`)
               if (!names.has(state) && !requiredContextStates.has(state)) continue
-              const localSetter = names.has(setter) || requiredContextStates.has(state) ? setter : `__kContextState_${state}`
-              setters.set(localSetter, state)
-              registerState(owner, state, localSetter, "context", node, { owner: hook.stateOwner, state: hook.stateSymbols.get(state) })
+              const privateFields = customHookPrivateFields.get(node) ?? []
+              const localName = field => {
+                if (names.has(field)) return field
+                const occupied = name => owner.parameters.some(parameter => bindingNames(parameter.name).includes(name)) || owner.body.statements.some(statement => statement !== node.parent.parent && statementDeclaresName(statement, name))
+                if (!occupied(field)) return field
+                let index = 0
+                let local
+                do local = `__kContext_${field}${index++ || ""}`
+                while (occupied(local))
+                return local
+              }
+              const localState = localName(state)
+              const localSetter = localName(setter)
+              setters.set(localSetter, localState)
+              registerState(owner, localState, localSetter, "context", node, { owner: hook.stateOwner, state: hook.stateSymbols.get(state) })
               if (requiredContextStates.has(state)) {
-                const fields = customHookPrivateFields.get(node) ?? []
-                for (const field of [state, setter]) {
-                  if (names.has(field) || fields.includes(field)) continue
-                  const conflict = owner.parameters.some(parameter => bindingNames(parameter.name).includes(field)) || owner.body.statements.some(statement => statement !== node.parent.parent && statementDeclaresName(statement, field))
-                  if (conflict) throw sourceNodeError(node.name, sourceFile, `Context action state field ${JSON.stringify(field)} conflicts with a consumer binding`)
-                  fields.push(field)
+                for (const [field, local] of [[state, localState], [setter, localSetter]]) {
+                  if (names.has(field) || privateFields.some(entry => (typeof entry === "string" ? entry : entry.property) === field)) continue
+                  privateFields.push({ property: field, local })
+                  if (field !== local) contextSubstitutions.set(field, factory.createIdentifier(local))
                 }
-                customHookPrivateFields.set(node, fields)
+                customHookPrivateFields.set(node, privateFields)
               }
               continue
             }
@@ -584,12 +618,14 @@ function createKudzuTransformer({ semantic, handlerUrl, file, sourceFiles, sourc
           settersByFunction.set(owner, setters)
           for (const name of names) {
             if (hook.callbacks.has(name)) {
+              const callback = contextSubstitutions.size ? substituteClone(hook.callbacks.get(name), contextSubstitutions, factory, context) : hook.callbacks.get(name)
               const callbacks = customHookFunctionsByOwner.get(owner) ?? new Map()
-              callbacks.set(name, hook.callbacks.get(name))
+              callbacks.set(name, callback)
               customHookFunctionsByOwner.set(owner, callbacks)
               if (hook.context) {
                 const reducers = reducersByFunction.get(owner) ?? new Map()
-                reducers.set(name, { contextAction: hook.callbacks.get(name), states: hook.states })
+                const states = new Map([...hook.states].map(([setter, state]) => [contextSubstitutions.get(setter)?.text ?? setter, contextSubstitutions.get(state)?.text ?? state]))
+                reducers.set(name, { contextAction: callback, states })
                 reducersByFunction.set(owner, reducers)
               }
             }
@@ -1549,7 +1585,9 @@ function createKudzuTransformer({ semantic, handlerUrl, file, sourceFiles, sourc
         const privateFields = customHookPrivateFields.get(node)
         return factory.updateVariableDeclaration(node, factory.updateObjectBindingPattern(node.name, [
           ...node.name.elements,
-          ...privateFields.map(name => factory.createBindingElement(undefined, undefined, name))
+          ...privateFields.map(entry => typeof entry === "string"
+            ? factory.createBindingElement(undefined, undefined, entry)
+            : factory.createBindingElement(undefined, entry.property === entry.local ? undefined : entry.property, entry.local))
         ]), node.exclamationToken, node.type, node.initializer)
       }
       if (ts.isBlock(node) && setterHookHelpers.has(node)) {

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@kudzujs/core",
-  "version": "0.8.43",
+  "version": "0.8.45",
   "description": "HTML-first TSX framework with synchronous state semantics and no virtual DOM",
   "type": "module",
   "license": "MIT",
@@ -60,6 +60,7 @@
     "benchmark:commerce": "node test/commerce-build-performance.mjs",
     "benchmark:module-cache": "node test/module-cache-performance.mjs",
     "benchmark:native": "node test/native-performance.mjs",
+    "benchmark:source-scale": "node test/source-scale-performance.mjs",
     "prepublishOnly": "npm run check && npm test",
     "deploy": "wrangler deploy",
     "preview": "wrangler dev"