feat(worker-bus): Phase 3 — Rust TaskSupervisor + text.analyze native task by qnbs · Pull Request #70 · qnbs/WorldScript-Studio

qnbs · 2026-06-02T23:21:39Z

WorkerBus v2 — Phase 3 (Rust TaskSupervisor) + build unblock + docs

Completes the native half of the hybrid router, unblocks the long-red Tauri build, and reconciles the docs.

Phase 3 — Rust TaskSupervisor

src-tauri/src/commands/task_supervisor.rs (new) + commands/mod.rs + lib.rs registration — storycraft_task_supervisor_ping (version) + storycraft_task_supervisor_submit (taskType dispatcher; unknown/bad payload → { success:false, error }, never a hard Err). First native task text.analyze (word/char/sentence/syllable + Flesch Reading Ease, pure Rust, 8 #[cfg(test)] tests).
services/rustTaskSupervisor.ts (new) — analyzeTextViaRust() probes isRustComputeAvailable() before routing (Rust-only task never hits the web pool; null → JS fallback); 5 unit tests.

Fixed — Tauri build (red since 2026-05-30, root-caused via CI dispatch)

src-tauri/Cargo.toml — removed unused, unresolvable specta = "2" / tauri-specta = "2" (only 2.0.0-rc.* exist → cargo resolution failed before any compile).
src-tauri/src/lora.rs — LoraEnvReport deserialized but derived only Serialize (E0277) → added Deserialize.
task_supervisor.rs — json import → test mod; #[allow(dead_code)] on wire-contract fields.

Docs

CHANGELOG, AUDIT, ADR 0003 (WorkerBus hybrid routing), docs/TAURI-CI (native-verification method + known blockers), AGENTS, README doc-hub, ROADMAP, TODO.

Verification

✅ Local: biome + tsc --noEmit + 5 TS tests green.
✅ Native (the only Rust gate — no PR CI compiles src-tauri/):tauri-build.yml dispatched on this branch now compiles clean (Finished release ~4m18s) and bundles .deb / .rpm / .AppImage on ubuntu. The run's final updater-signing step fails on a malformed TAURI_SIGNING_PRIVATE_KEY repo secret (maintainer task), and the Windows runner fails in the setup composite (infra) — both unrelated to this code.

🤖 Generated with Claude Code

… task Completes the WorkerBus v2 Rust half. The TS bridge (services/tauriTaskBridge.ts + hybridRouter.ts) already invokes storycraft_task_supervisor_{submit,ping}; this adds the missing native commands. - src-tauri/src/commands/task_supervisor.rs: ping (version) + submit dispatcher. Unknown/bad-payload tasks resolve as { success:false, error } (never a hard Err), matching the RustTaskResultEvent honest-failure convention so the router's fallback is driven by result.success. First real task `text.analyze` computes word/char/sentence/syllable counts + Flesch Reading Ease (pure Rust, no new deps). 8 #[cfg(test)] unit tests cover analysis + dispatch. - commands/mod.rs + lib.rs: register both commands in generate_handler!. - services/rustTaskSupervisor.ts: analyzeTextViaRust() probes isRustComputeAvailable() before routing so a Rust-only task never hits the web worker pool; returns null (JS fallback) when unavailable. 5 unit tests. Verified locally: biome + tsc --noEmit + the 5 TS tests green. Rust cannot be compiled on this low-end host (pre-existing specta="^2" resolution issue, unrelated to this change) — gated by tauri-build.yml (workflow_dispatch), triggered on push. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>

codeant-ai · 2026-06-02T23:21:44Z

CodeAnt AI is reviewing your PR.

Thanks for using CodeAnt! 🎉

We're free for open-source projects. if you're enjoying it, help us grow by sharing.

Share on X ·
Reddit ·
LinkedIn

vercel · 2026-06-02T23:21:45Z

The latest updates on your projects. Learn more about Vercel for GitHub.

Project	Deployment	Actions	Updated (UTC)
storycraft-studio	Ready	Preview, Comment	Jun 2, 2026 11:53pm

codeant-ai · 2026-06-02T23:23:10Z

+
+ let outcome: Result<Value, String> = match request.task_type.as_str() {
+ "text.analyze" => run_text_analyze(&request.payload),
+ other => Err(format!("Unknown task type: {other}")),


Suggestion: Return a generic client-safe error message for unsupported task types and log the specific task type only in internal diagnostics. [custom_rule_security]
Severity Level: Critical 🚨
Why it matters? 🤔
The code returns a user-visible error string that includes the unsupported task type. That is a real security-style information leak from the external API surface, so the suggestion is correctly identifying a violation.
Fix in Cursor | Fix in VSCode Claude
(Use Cmd/Ctrl + Click for best experience)
Prompt for AI Agent 🤖
This is a comment left during a code review. **Path:** src-tauri/src/commands/task_supervisor.rs **Line:** 86:86 **Comment:***Custom Rule Security: Return a generic client-safe error message for unsupported task types and log the specific task type only in internal diagnostics. Validate the correctness of the flagged issue. If correct, How can I resolve this? If you propose a fix, implement it and please make it concise. Once fix is implemented, also check other comments on the same PR, and ask user if the user wants to fix the rest of the comments as well. if said yes, then fetch all the comments validate the correctness and implement a minimal fix
👍 | 👎

codeant-ai · 2026-06-02T23:23:10Z

+ let text = payload
+ .get("text")
+ .and_then(Value::as_str)
+ .ok_or_else(|| "text.analyze requires payload.text to be a string".to_string())?;


Suggestion: Replace schema-specific validation text with a neutral user-facing error message, and keep payload shape details only in internal logs. [custom_rule_security]
Severity Level: Critical 🚨
Why it matters? 🤔
The validation error is returned to the caller and explicitly reveals the internal payload shape requirement (payload.text). This is a direct exposure of contract details, so the suggestion matches a real violation.
Fix in Cursor | Fix in VSCode Claude
(Use Cmd/Ctrl + Click for best experience)
Prompt for AI Agent 🤖
This is a comment left during a code review. **Path:** src-tauri/src/commands/task_supervisor.rs **Line:** 111:114 **Comment:***Custom Rule Security: Replace schema-specific validation text with a neutral user-facing error message, and keep payload shape details only in internal logs. Validate the correctness of the flagged issue. If correct, How can I resolve this? If you propose a fix, implement it and please make it concise. Once fix is implemented, also check other comments on the same PR, and ask user if the user wants to fix the rest of the comments as well. if said yes, then fetch all the comments validate the correctness and implement a minimal fix
👍 | 👎

`specta = "2"` / `tauri-specta = "2"` are referenced in no .rs file (build.rs only calls tauri_build::build()) and are absent from Cargo.lock. The `"2"` (=^2 stable) requirement matches nothing on crates.io — only `2.0.0-rc.*` prereleases exist — so `cargo build` fails at resolution before compiling anything: error: failed to select a version for the requirement `specta = "^2"` This is why tauri-build.yml has been red since 2026-05-30 (the older lora.rs errors were already past this gate). Removing the two dead deps is the root-cause fix: resolution succeeds, the whole Rust crate (incl. the Phase 3 task_supervisor) can build, and nothing functional is lost. TS-binding generation can re-add tauri-specta later with a working exact rc pin. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>

codeant-ai · 2026-06-02T23:27:30Z

+ const handle = await routeTask<RustTextAnalysis>(
+ 'text.analyze',
+ { text },
+ { target: 'rust', rustComputeEnabled: true, priority: 'low' },
+ );


Suggestion: This call still goes through routeTask, which is designed to fall back to the web worker pool on Rust failure. That breaks the intended "Rust-only" routing guarantee and can enqueue text.analyze on web workers during transient Rust errors; call the Rust bridge directly (or add a strict no-fallback mode) for this path. [logic error]
Severity Level: Major ⚠️
- ⚠️ Rust text.analyze may run on web worker fallback. - ⚠️ Rust-only compute flag cannot enforce native-only execution.
Steps of Reproduction ✅
1. Call `analyzeTextViaRust()` from `services/rustTaskSupervisor.ts:33-52` with a non-empty string and `{ rustComputeEnabled: true }` so it proceeds past the early null returns. 2. Inside `analyzeTextViaRust()`, `isRustComputeAvailable()` from `services/tauriTaskBridge.ts:43-57` is awaited; assume it resolves `true`, then the helper calls `routeTask<RustTextAnalysis>('text.analyze', {text}, {target: 'rust', rustComputeEnabled: true, priority: 'low'})` at `services/rustTaskSupervisor.ts:41-45`. 3.`routeTask()` in `services/hybridRouter.ts:36-87` enters the Rust branch (lines 43-77), calls `invokeRustTask(request)` at lines 48-56, and wraps it in a `try {...} catch (err) {...}` block; if `invokeRustTask()` throws (for example, due to a transient Tauri error or missing Rust command), the catch at lines 72-75 logs `'Rust route failed — falling back to web worker pool'` and deliberately falls through to the web worker path. 4. The web worker path at `services/hybridRouter.ts:79-87` enqueues the same `taskType` and payload via `bus.enqueue(taskType, payload, { ...busOpts, target: 'web' })`, returns a `TaskHandle`, and `analyzeTextViaRust()` at `services/rustTaskSupervisor.ts:46-47` awaits `handle.result`, so the supposedly Rust-only `'text.analyze'` task can be executed on the WorkerBus web pool whenever the Rust invoke fails, contradicting the comment on lines 25-31 that this helper should never enqueue onto the web worker pool.
Fix in Cursor | Fix in VSCode Claude
(Use Cmd/Ctrl + Click for best experience)
Prompt for AI Agent 🤖
This is a comment left during a code review. **Path:** services/rustTaskSupervisor.ts **Line:** 41:45 **Comment:***Logic Error: This call still goes through `routeTask`, which is designed to fall back to the web worker pool on Rust failure. That breaks the intended "Rust-only" routing guarantee and can enqueue `text.analyze` on web workers during transient Rust errors; call the Rust bridge directly (or add a strict no-fallback mode) for this path. Validate the correctness of the flagged issue. If correct, How can I resolve this? If you propose a fix, implement it and please make it concise. Once fix is implemented, also check other comments on the same PR, and ask user if the user wants to fix the rest of the comments as well. if said yes, then fetch all the comments validate the correctness and implement a minimal fix
👍 | 👎

codeant-ai · 2026-06-02T23:27:30Z

+ pub payload: Value,
+ pub priority: String,
+ pub target: String,
+ pub timeout_ms: u64,


Suggestion: The request includes timeout_ms, but task execution ignores it and runs handlers directly, so long-running or stuck tasks will never time out despite the contract exposing a timeout field. Enforce the timeout in the dispatcher so callers get bounded execution behavior. [incomplete implementation]
Severity Level: Major ⚠️
- ⚠️ Rust TaskSupervisor ignores timeoutMs for all tasks. - ⚠️ Callers cannot bound execution time despite timeout field.
Steps of Reproduction ✅
1. A TypeScript caller invokes `routeTask()` from `services/hybridRouter.ts:36-87` with options including `timeoutMs`, for example `routeTask('text.analyze', payload, {target: 'rust', rustComputeEnabled: true, timeoutMs: 1000})`. 2. In `routeTask()`, the Rust request envelope is built at `services/hybridRouter.ts:48-55` as `const request: RustTaskRequest = {taskId, taskType, payload, priority, target: 'rust', timeoutMs: opts.timeoutMs??300_000};`, so the configured timeout is serialized into the `RustTaskRequest`. 3.`invokeRustTask()` in `services/tauriTaskBridge.ts:15-23` forwards this `RustTaskRequest` over Tauri IPC to the Rust command `storycraft_task_supervisor_submit`, which is registered in `src-tauri/src/lib.rs:52-61` and implemented in `src-tauri/src/commands/task_supervisor.rs:80-107`. 4. Inside `storycraft_task_supervisor_submit`, the `RustTaskRequest` struct defined at `src-tauri/src/commands/task_supervisor.rs:29-38` includes `pub timeout_ms: u64`, but the function body only inspects `request.task_type` (lines 84-87) and never reads `request.timeout_ms`; a repo-wide Grep for `timeout_ms` shows usage only in the struct definition and test fixtures (lines 261, 278, 296), so task handlers (currently `run_text_analyze`) always run to completion with no enforcement of the caller-provided timeout.
Fix in Cursor | Fix in VSCode Claude
(Use Cmd/Ctrl + Click for best experience)
Prompt for AI Agent 🤖
This is a comment left during a code review. **Path:** src-tauri/src/commands/task_supervisor.rs **Line:** 35:35 **Comment:***Incomplete Implementation: The request includes `timeout_ms`, but task execution ignores it and runs handlers directly, so long-running or stuck tasks will never time out despite the contract exposing a timeout field. Enforce the timeout in the dispatcher so callers get bounded execution behavior. Validate the correctness of the flagged issue. If correct, How can I resolve this? If you propose a fix, implement it and please make it concise. Once fix is implemented, also check other comments on the same PR, and ask user if the user wants to fix the rest of the comments as well. if said yes, then fetch all the comments validate the correctness and implement a minimal fix
👍 | 👎

codeant-ai · 2026-06-02T23:27:30Z

+ let words: Vec<&str> = text.split_whitespace().filter(|w| !w.is_empty()).collect();
+ let word_count = words.len();
+
+ let sentence_count = count_sentences(text);
+ let syllable_count: usize = words.iter().map(|w| count_syllables(w)).sum();


Suggestion: Building a full Vec<&str> of all words before counting syllables adds avoidable memory pressure for large manuscripts. This task is explicitly for large inputs, so compute word and syllable counts in a streaming pass instead of collecting all tokens first. [performance]
Severity Level: Major ⚠️
- ⚠️ Text.analyze allocates Vec for every word token. - ⚠️ Large manuscripts increase memory footprint unnecessarily.
Steps of Reproduction ✅
1. When a `text.analyze` task is submitted, `storycraft_task_supervisor_submit()` at `src-tauri/src/commands/task_supervisor.rs:80-107` dispatches to `run_text_analyze()` for `task_type == "text.analyze"` (lines 84-86), and `run_text_analyze()` at lines 110-118 extracts `payload["text"]` then calls `analyze_text(text)`. 2.`analyze_text()` in `src-tauri/src/commands/task_supervisor.rs:122-151` computes character counts, then allocates `let words: Vec<&str>=text.split_whitespace().filter(|w|!w.is_empty()).collect();`atline126andderives`letword_count=words.len();`atline127,storingaborrowedsliceforeverytoken.3.Thefunctionthenseparatelycomputes`letsentence_count=count_sentences(text);`atline129,whichre-iteratesovertheentire`text`,and`letsyllable_count:usize=words.iter().map(|w|count_syllables(w)).sum();`atline130,traversingthe`words`vectortocountsyllablespertoken.4.Forlargemanuscripts(themodulecommentatlines7-9explicitlycallsoutoffloading"large manuscripts"),eachcallto`analyze_text()`materializesa`Vec`ofallwhitespace-separatedtokenseventhoughonlyaggregatecountsareused;thisper-wordallocationincreasesmemoryandallocationoverheadonevery`text.analyze`task,whereasastreamingpassover`split_whitespace()`couldcomputeboth`word_count`and`syllable_count`withoutstoringthefulllist,matchingthe"no allocation beyond theword iterator"intentstatedatlines120-121.
Fix in Cursor | Fix in VSCode Claude
(Use Cmd/Ctrl + Click for best experience)
Prompt for AI Agent 🤖
This is a comment left during a code review. **Path:** src-tauri/src/commands/task_supervisor.rs **Line:** 126:130 **Comment:***Performance: Building a full `Vec<&str>` of all words before counting syllables adds avoidable memory pressure for large manuscripts. This task is explicitly for large inputs, so compute word and syllable counts in a streaming pass instead of collecting all tokens first. Validate the correctness of the flagged issue. If correct, How can I resolve this? If you propose a fix, implement it and please make it concise. Once fix is implemented, also check other comments on the same PR, and ask user if the user wants to fix the rest of the comments as well. if said yes, then fetch all the comments validate the correctness and implement a minimal fix
👍 | 👎

codeant-ai · 2026-06-02T23:27:35Z

CodeAnt AI finished reviewing your PR.

…port Two fixes surfaced once the specta resolution block was removed and the Rust crate actually compiled on CI: - lora.rs: `check_lora_environment` deserializes the Python sidecar's JSON stdout via `serde_json::from_str::<LoraEnvReport>` (lora.rs:209), but the struct derived only Serialize -> E0277 broke the whole-crate compile (the real reason tauri-build.yml has been red since 2026-05-30). Add Deserialize (already imported). - task_supervisor.rs: `json` is only used in #[cfg(test)] -> move the import into the test module to silence the unused-import warning in the non-test build. Verified: this is the last error/warning the ubuntu+macos runners reported after the specta fix; re-dispatching tauri-build to confirm a clean native compile. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>

…t fields The native ubuntu/macos build now compiles cleanly and produces .deb/.rpm/.AppImage bundles (verified via tauri-build dispatch). The only remaining diagnostic was a dead_code warning: priority/target/timeoutMs/retryPolicy are accepted from the TS router per the worker-bus contract but not yet read by the dispatcher. allow(dead_code) with a note until the retry/timeout path is implemented. NOTE: tauri-build still exits non-zero at the very end on the Tauri updater signing step ("incorrect updater private key password: Missing comment in secret key") — a malformed repo signing secret, unrelated to code; the app + all 3 bundles build fine. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>

…d findings Bring the living/canonical docs in line with the current code + CI reality. Sprint handoffs, docs/history, and version-sprint docs are archival (point-in-time) and left frozen by design. - CHANGELOG.md: 2026-06-03 cluster — WorkerBus v2 Phase 3 (Rust TaskSupervisor + text.analyze), Local-AI Phase 2.3 perf sprint (pipelineLruCache dispose-on-evict, aiRetry property tests, kokoro, ADRs, coverage ratchet L74/B60/F66/S72), and the Tauri-build unblock (### Fixed). - AUDIT.md: new 2026-06-03 audit section (Phase 3 + the 3 root-caused tauri-build blockers + the remaining non-code updater-signing / Windows-setup issues); follow-up chain extended. - docs/adr/0003-workerbus-hybrid-routing.md (new) + ADR index: records the v2 runtime → manager → hybrid router → legacy adapter → Rust TaskSupervisor layering, the honest-failure contract, and the "no PR-CI gate for Rust" verification constraint. - docs/TAURI-CI.md: "Verifying native (Rust) changes" (dispatch-on-branch is the gate) + "Build health & known blockers (2026-06-03)". - AGENTS.md: agent rule — no PR-CI gate for src-tauri; dispatch tauri-build on the branch. - README.md: ADR doc-hub entry now lists WorkerBus hybrid routing. - ROADMAP.md: Local-AI Perfection Phases 2.3/2.4 marked done; WorkerBus Phase 3 noted. No i18n keys added (docs only); badges left as-is (match last CI-measured numbers). Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>

codeant-aiBot added the size:L This PR changes 100-499 lines, ignoring generated files label Jun 2, 2026

codeant-aiBot reviewed Jun 2, 2026
View reviewed changes

qnbsand others added 3 commits June 3, 2026 01:34

qnbs merged commit eff15c1 into mainJun 3, 2026
16 checks passed

qnbs deleted the feat/workerbus-v2-phase3-rust branch June 3, 2026 00:11

qodo-code-reviewBot mentioned this pull request Aug 18, 2026
refactor(desktop): introduce renderer-neutral platform contracts #384
Merged
4 tasks

Navigation Menu

feat(worker-bus): Phase 3 — Rust TaskSupervisor + text.analyze native task - #70

feat(worker-bus): Phase 3 — Rust TaskSupervisor + text.analyze native task#70
qnbs merged 5 commits into
mainfrom
feat/workerbus-v2-phase3-rust

qnbs commented Jun 2, 2026•
edited
Loading

Uh oh!

codeant-aiBot commented Jun 2, 2026

Uh oh!

vercelBot commented Jun 2, 2026•
edited
Loading

Uh oh!

codeant-aiBotJun 2, 2026

Uh oh!

codeant-aiBotJun 2, 2026

Uh oh!

Uh oh!

codeant-aiBotJun 2, 2026

Uh oh!

codeant-aiBotJun 2, 2026

Uh oh!

codeant-aiBotJun 2, 2026

Uh oh!

codeant-aiBot commented Jun 2, 2026

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

1 participant

Conversation

qnbs commented Jun 2, 2026• edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

WorkerBus v2 — Phase 3 (Rust TaskSupervisor) + build unblock + docs

Phase 3 — Rust TaskSupervisor

Fixed — Tauri build (red since 2026-05-30, root-caused via CI dispatch)

Docs

Verification

Uh oh!

codeant-aiBot commented Jun 2, 2026

Thanks for using CodeAnt! 🎉

Uh oh!

vercelBot commented Jun 2, 2026• edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

codeant-aiBotJun 2, 2026

Choose a reason for hiding this comment

Uh oh!

codeant-aiBotJun 2, 2026

Choose a reason for hiding this comment

Uh oh!

Uh oh!

codeant-aiBotJun 2, 2026

Choose a reason for hiding this comment

Uh oh!

codeant-aiBotJun 2, 2026

Choose a reason for hiding this comment

Uh oh!

codeant-aiBotJun 2, 2026

Choose a reason for hiding this comment

Uh oh!

codeant-aiBot commented Jun 2, 2026

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

1 participant

qnbs commented Jun 2, 2026•
edited
Loading

vercelBot commented Jun 2, 2026•
edited
Loading