autoworker
Auto-loop execution workflow with quality gates. Use when starting any non-trivial implementation task. Provides automatic task decomposition, code…
Complete subtask verification plan: upstream traceability table + L1-L4 test plan + self-check. Call after autoworker:subtask-init. Makes subtask ready for code implementation.
$ npx -y skills add phj128/autoworker --skill subtask-plan --agent claude-codeHow it fires
How this skill gets triggered: by you, by Claude, or both.
/subtask-planContext preview
The summary Claude sees to decide when to auto-load this skill.
Complete subtask verification plan: upstream traceability table + L1-L4 test plan + self-check. Call after autoworker:subtask-init. Makes subtask ready for code implementation.
name: subtask-plan description: | Complete subtask verification plan: upstream traceability table + L1-L4 test plan + self-check. Call after autoworker:subtask-init. Makes subtask ready for code implementation.
Invoked after autoworker:subtask-init completes. Fills in upstream traceability table, L1-L4 verification plan, and solution self-check.
Glob `subtask_*.md` (exclude subtask_template.md) →
0 found → stop, prompt to create subtask
1 found → use directly (backward compatible)
multiple → grep `status:` to filter:
- Files without status field treated as active (backward compatible)
- Exactly 1 active → use it
- 0 active → list all files + status, prompt user to choose
- >1 active → report anomaly
→ Read → check first-half completeness**Pre-checks**:
**Source**: Read plan file (if exists) + review conversation for user's verification/testing requirements.
Edit subtask's "Upstream Verification Traceability" table:
| Upstream text (copy verbatim) | Subtask item | Delta rationale | |------------------------------|--------------|-----------------|
**Rules**:
Read the subtask's "Acceptance Criteria" table, for each metric ask:
> If this metric deviates, which L1-L4 layer would catch it?
**Purpose**: Prevent acceptance criteria defined in the plan but not covered by the verification plan (silent failure).
Cross-reference the traceability table, Edit subtask to fill in each layer's verification plan:
**L1 Build**:
**L2 Unit** (can skip but state reason):
**L3 Chain** (can skip but state reason):
**L4 End-to-End** (mandatory):
**Verification coverage table**: | Modified file | Change content | Corresponding verification item | Each file planned for modification must have a corresponding verification item.
**Constraints**:
Build metric-to-test mapping table:
| Acceptance metric | Corresponding test item (which L, which item) | Coverage status | |------------------|-----------------------------------------------|----------------|
Answer each one (any unsatisfactory → report the issue, do not continue):
0. **Is the direction right?** Derived from problem essence, or pattern-matched from similar scenarios? 1. **Is the abstraction level right?** Does the caller naturally possess this capability? 2. **10x robustness**: Still robust if input diversity increases 10x? 3. **Does a simpler solution exist?** Same effect, less code/assumptions?
Edit self-check results into subtask's progress log.
Subtask ready: subtask_<sequence>_<name>.md - Traceability: <N> upstream requirements aligned - Verification plan: L1 done, L2 <done/skip>, L3 <done/skip>, L4 done - Coverage: <N>/<N> files have verification items - Acceptance criteria coverage: <N>/<N> metrics have corresponding tests - Self-check: 4/4 passed → Automatically invoking autoworker:dispatch
**After outputting the summary above, immediately invoke `autoworker:dispatch`. Do not wait for user instructions, do nothing else.**
An auto-loop execution workflow with quality gates for Claude Code. Give Claude a task. Autoworker decomposes it, implements code, runs tests, and iterates through quality gates — autonomously looping until the job is done right.
Auto-loop execution workflow with quality gates. Use when starting any non-trivial implementation task. Provides automatic task decomposition, code…
Record progress to subtask: Phase completion (from autoworker:code) or test results (from autoworker:test). Auto-detects upstream type from conversation…
Implement ONE pending Phase from subtask plan. Only write code, do not mark checkboxes or run tests. Called by autoworker:dispatch. Ends by calling…
Structured deep discussion for Plan Mode. Runs inside EnterPlanMode to ensure thorough questioning before writing a plan. Covers motivation, assumptions,…
Read subtask.md checkbox state and route to next skill. The ONLY routing point in the execution loop. Called after autoworker:checkpoint,…
Final quality gate before reporting task completion (Gate 3). Fills confidence assessment, runs supplementary verification for <95% items, completes self-check…