skills/orchestrating-workflows/SKILL.md
Workflow orchestration for /code, /fix, /audit. Use when: /code ワークフロー, /fix ワークフロー, quality gates, 品質ゲート, RGRC サイクル.
npx skillsauth add thkt/dotclaude orchestrating-workflowsInstall this skill globally with one command. Works with Claude Code, Cursor, and Windsurf.
3 of 9 scanners reported clean
Some scanners were skipped, did not run, or reported a non-clean status. Review each row below.
| Command | Workflow Reference |
| ------- | ------------------------------------------------- |
| /code | ${CLAUDE_SKILL_DIR}/references/code-workflow.md |
| /fix | ${CLAUDE_SKILL_DIR}/references/fix-workflow.md |
| Pattern | Reference |
| --------------- | ---------------------------------------------------------------------------- |
| IDR Generation | hooks/lifecycle/idr-pre-commit.sh |
| TDD Cycle | ${CLAUDE_SKILL_DIR}/references/tdd-cycle.md |
| Test Generation | ${CLAUDE_SKILL_DIR}/references/test-generation.md |
| Gate | Target | Verification |
| ------------ | ---------------- | ------------------------------------------ |
| Tests | All passing | npm test exit code 0 |
| Lint | 0 errors | npm run lint exit code 0 |
| Types | No errors | tsc --noEmit exit code 0 |
| Coverage | C0 ≥90%, C1 ≥80% | Coverage report |
| Test Quality | ≥70 | test-quality-evaluator (skip if no Spec) |
When a Spec with Test Scenarios exists, spawn test-quality-evaluator as a
background agent:
Agent(subagent_type: "test-quality-evaluator",
prompt: "spec_path: <path>\ntest_paths: <paths>",
run_in_background: true)
Score ≥70 → pass. Score <70 → report uncovered/excess/intent issues, fix before
proceeding. Skip when no Spec exists (e.g., /fix, ad-hoc changes).
After RGRC cycles, spawn code-quality-reviewer as a background agent:
Agent(subagent_type: "code-quality-reviewer",
prompt: "Review files changed in this session: <paths>",
run_in_background: true)
High severity → fix before Quality Gates. Medium/low → advisory (note in IDR).
Skip for /fix and single-file changes.
Tests: pass | fail (detail)
Lint: pass | fail (detail)
Types: pass | fail (detail)
Coverage: C0 XX% / C1 XX% — pass | fail
Test Quality: XX/100 — pass | skip (no Spec)
All 5 lines required. Empty lines indicate a skipped gate — investigate before proceeding.
| Excuse | Counter |
| ---------------------------------------- | -------------------------------------------------------------- |
| "Tests pass, lint can wait" | Lint errors are tech debt. Zero errors before commit |
| "Type errors are just warnings" | tsc --noEmit exit 0 or no ship. Type warnings are errors |
| "Coverage is close enough" | "Close enough" is failure with extra steps. Meet the threshold |
| "This gate doesn't apply to this change" | All 4 gates apply to every change. No exceptions |
tools
Delegate implementation to codex (coder) via the herdr-agentchat plugin and drive a two-pane conversation to completion.
development
Extract recurring patterns from past closed PRs/issues and the research findings in workspace/research/, verify them against the latest code, and propose them to docs/wiki/ via PR.
development
Create Decision Records (DR) in MADR v4 format with auto-numbering.
development
Detect flaky tests by shaking them — repeated runs under varied order, parallelism, and seed — plus a static smell scan that flags latent flakiness in tests that currently pass. Classify each target as confirmed-flaky, latent-flaky, or stable and fix the root cause without weakening the test. Do NOT use to fix a confirmed single bug (use /fix) or for static-only code review (use /audit).