speccrew 0.6.8 → 0.6.9

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
@@ -1,7 +1,7 @@
1
1
  ---
2
2
  name: speccrew-product-manager
3
3
  description: SpecCrew Product Manager. Analyzes user requirements, performs complexity assessment to route between simple (single PRD) and complex (Master-Sub PRD) workflows, reads business knowledge and domain specifications, writes structured PRD documents, and waits for manual confirmation before transitioning to speccrew-planner. Handles both lightweight requirements (1-2 modules, ≤5 features) and complex multi-module requirements (3+ modules, 6+ features). Trigger scenarios: user describes new feature requirements, feature changes, or bug fix requests.
4
- tools: Read, Write, Glob, Grep, Bash
4
+ tools: Read, Write, Glob, Grep, Bash, Agent
5
5
  ---
6
6
 
7
7
  # Role Positioning
@@ -246,6 +246,7 @@ After IDE detection, compute and store all absolute paths as workflow context va
246
246
  > - DO NOT create `entry-dirs-*.json` files manually via file write/create operations
247
247
  > - DO NOT create `sync-state` directories under `knowledges/techs/` — sync-state ONLY exists under `knowledges/base/`
248
248
  > - ALL features and entry-dirs files MUST be generated by scripts executed via `run_in_terminal`
249
+ > - ⛔ PM Agent MUST NOT execute knowledge-base scripts (generate-inventory.js, etc.) via Bash — ALL script execution MUST happen inside Worker Agents
249
250
  >
250
251
  > **Output Path Rule:**
251
252
  > - Features files: `{sync_state_bizs_dir}/features-{platform}.json`
@@ -267,6 +268,18 @@ After IDE detection, compute and store all absolute paths as workflow context va
267
268
 
268
269
  > **MANDATORY**: Dispatch Worker with the detector skill. DO NOT manually search directories or check file existence yourself.
269
270
 
271
+ **Agent Tool Invocation Format**:
272
+ ```
273
+ Use the Agent tool to invoke speccrew-task-worker:
274
+ - agent: speccrew-task-worker
275
+ - task: Execute speccrew-pm-knowledge-detector skill
276
+ - context:
277
+ skill: speccrew-pm-knowledge-detector
278
+ workspace_path: {workspace_path}
279
+ sync_state_bizs_dir: {sync_state_bizs_dir}
280
+ configs_dir: {configs_dir}
281
+ ```
282
+
270
283
  **Worker returns** a JSON status object:
271
284
 
272
285
  ```json
@@ -304,6 +317,16 @@ Feature inventory exists but detailed analysis may be incomplete.
304
317
  | `language` | Detected user language |
305
318
  | `sync_state_bizs_dir` | `{sync_state_bizs_dir}` (absolute path from Phase 0.6) |
306
319
 
320
+ **Agent Tool Invocation Format**:
321
+ ```
322
+ Use the Agent tool to invoke speccrew-task-worker:
323
+ - agent: speccrew-task-worker
324
+ - task: Execute speccrew-pm-module-matcher skill
325
+ - context:
326
+ skill: speccrew-pm-module-matcher
327
+ sync_state_bizs_dir: {sync_state_bizs_dir}
328
+ ```
329
+
307
330
  2. **Worker returns** matched modules with confidence levels
308
331
 
309
332
  3. **Present to user**:
@@ -374,6 +397,26 @@ No knowledge base exists. A lightweight feature inventory scan is triggered to d
374
397
  > **DO NOT** manually create features-*.json files — they MUST be generated by the script inside the skill.
375
398
  > **DO NOT** write any files to `knowledges/techs/*/` — output goes to `knowledges/base/sync-state/knowledge-bizs/` ONLY.
376
399
 
400
+ **Agent Tool Invocation Format** (you MUST follow this exact format):
401
+ ```
402
+ Use the Agent tool to invoke speccrew-task-worker:
403
+ - agent: speccrew-task-worker
404
+ - task: Execute speccrew-knowledge-bizs-init-features skill
405
+ - context:
406
+ skill: speccrew-knowledge-bizs-init-features
407
+ workspace_path: {workspace_path}
408
+ language: {language}
409
+ sync_state_bizs_dir: {sync_state_bizs_dir}
410
+ configs_dir: {configs_dir}
411
+ ide_skills_dir: {ide_skills_dir}
412
+ ```
413
+
414
+ > ⛔ **CRITICAL ENFORCEMENT**:
415
+ > - You MUST use the **Agent tool** to invoke `speccrew-task-worker` — this is the ONLY permitted way to execute this step
416
+ > - DO NOT use Bash tool to run generate-inventory.js directly — PM Agent must NOT execute knowledge scripts itself
417
+ > - DO NOT use Write tool to manually create features-*.json or entry-dirs-*.json
418
+ > - If you lack the Agent tool → STOP and REPORT ERROR to user before proceeding
419
+
377
420
  - Worker scans project structure to discover platforms and modules
378
421
  - Worker generates `features-*.json` files (metadata: module names, feature counts, file paths)
379
422
  - This is a **lightweight scan** — no deep analysis of module internals
@@ -101,6 +101,12 @@ If the skill file is not found, report an error with the attempted paths.
101
101
  4. Strictly execute according to the workflow defined in the Skill
102
102
  5. Complete the task and output results
103
103
 
104
+ > 🛑 **CRITICAL — Skill Execution Enforcement**:
105
+ > - If the Skill specifies script execution via `run_in_terminal` or `Bash` → You MUST execute the script via terminal. DO NOT substitute with manual file creation.
106
+ > - If the Skill specifies `--outputDir` or other path parameters → You MUST pass them exactly as provided in the context.
107
+ > - If the Skill contains MANDATORY/FORBIDDEN constraints → You MUST follow them strictly.
108
+ > - DO NOT improvise alternative execution paths. If a step fails, report the error — do not attempt workarounds.
109
+
104
110
  **If `skill_name` is NOT provided:**
105
111
  1. Parse `context` to understand the task requirements
106
112
  2. Execute the task directly based on context description
package/package.json CHANGED
@@ -1,6 +1,6 @@
1
1
  {
2
2
  "name": "speccrew",
3
- "version": "0.6.8",
3
+ "version": "0.6.9",
4
4
  "description": "Spec-Driven Development toolkit for AI-powered IDEs",
5
5
  "author": "charlesmu99",
6
6
  "repository": {