メニュー
Run all done-criteria checks (CQ/T/CC/CL/S/I/FV/CV) and produce a pass/fail verification report. Use after implementing code, before submitting evidence to the quorum audit.
Codex または Claude でインストール この Prompt をコピーして Codex、Claude、または他のアシスタントに貼り付けると、Skill ページを確認してインストールできます。
SOC 職業分類に基づく
| name | quorum:verify |
| description | Run all done-criteria checks (CQ/T/CC/CL/S/I/FV/CV) and produce a pass/fail verification report. Use after implementing code, before submitting evidence to the quorum audit. |
| argument-hint | [optional: specific category - CQ, T, CC, CL, S, I, FV, CV] |
| model | claude-sonnet-4-6 |
| allowed-tools | Read, Grep, Glob, Bash(npx *), Bash(node *), Bash(python *), Bash(cargo *), Bash(go *), Bash(ruff *), Bash(git diff *), Bash(git status *), Bash(cat *), Bash(ls *) |
Run all done-criteria checks before evidence submission. Criteria from ${CLAUDE_PLUGIN_ROOT}/core/templates/references/${locale}/done-criteria.md.
| Context | Behavior |
|---|---|
| Interactive | Run checks → present report → suggest fixes |
| Headless | Run checks → output report → exit with status code (0 = all pass, 1 = failures) |
In headless mode, do NOT ask "should I fix this?" — output the report and exit. The caller (implementer agent) reads the report and decides.
| # | Category | Key Checks | Tool |
|---|---|---|---|
| 1 | Code Quality (CQ) | Per-file + project-wide checks from quality_rules.presets, audit-scan type-safety | Bash |
| 2 | Test (T) | Execute evidence test commands, check direct tests exist | Bash |
| 3 | Claim-Code (CC) | Diff scope vs Changed Files | Bash, Grep |
| 4 | Cross-Layer (CL) | BE→FE contracts, consumer existence | Read, Grep |
| 5 | Security (S) | Input validation, auth guards, audit-scan hardcoded | Grep, Read |
| 6 | i18n (I) | Locale keys in ALL locale files | Grep |
| 7 | Frontend (FV) | Page loads, DOM elements, console errors, build | Browser (if FE) |
| 8 | Coverage (CV) | stmt ≥ 85%, branch ≥ 75% per changed file | tool-runner.mjs |
consensus.trigger_tag, consensus.watch_filetrigger_tagNo trigger_tag section → "No evidence to verify" → stop.
Read references/checks.md for detailed commands and criteria per category.
## Verification Report
| # | Category | Status | Details |
|---|----------|--------|---------|
| 1 | Code Quality (CQ) | PASS / X issues | ... |
| 2 | Test (T) | PASS / X issues | ... |
| 3 | Claim-Code (CC) | PASS / X issues | ... |
| 4 | Cross-Layer (CL) | PASS / N/A | ... |
| 5 | Security (S) | PASS / X issues | ... |
| 6 | i18n (I) | PASS / X issues | ... |
| 7 | Frontend (FV) | PASS / SKIP | ... |
| 8 | Coverage (CV) | PASS / X issues | ... |
**Total: X/8 passed, Y issues found**
If all pass → "Ready for evidence submission." If any fail → list issues with fix recommendations.
Every category row (1-8) must have a status:
| Status | When |
|---|---|
| PASS | Check ran and passed |
| FAIL (N issues) | Check ran and found problems |
| SKIP | Not applicable — must include reason |
| N/A | Cannot be evaluated — must include reason |
A report with any blank status cell is incomplete. Fill all 8 rows before outputting.
Design tasks into tracks with work breakdowns and execution order. Also writes and maintains PRDs (Product Requirements Documents) — when a user requests a feature, analyzes it, adds it to the PRD, then decomposes into work breakdowns. Use for new feature planning, PRD writing, architecture changes, multi-track decomposition, requirements analysis, or adjusting existing execution plans. Trigger when user says things like 'add feature X', 'I need Y', 'plan Z', or describes any product requirement.
Run a consensus-loop audit manually — Codex reviews pending trigger_tag items in the watch file. Use when you want to trigger an audit without editing the watch file, re-run a failed audit, or test the audit prompt.
Squash-merge a worktree branch into the target branch with a structured commit message. Use after audit consensus and retrospective completion.
Session orchestrator for the consensus-loop — reads handoff, picks unblocked tasks, distributes to parallel workers, tracks agent assignments, manages correction cycles via SendMessage. Use when starting a work session with pending tasks, distributing implementation work, or reviewing completed worker output.
Extract learnings from audit history and conversation, manage memories, clean up stale entries. Use after completing a track, during retrospective (③ memory step), at end of session, or anytime the user wants memory maintenance. Triggers on 'what did we learn', 'memory cleanup', 'review learnings', 'retrospective', 'update memories', '회고', '메모리 정리'.
Show current consensus-loop status — pending reviews, audit state, retro marker, agent assignments. Use to check what's happening before starting work or after returning from a break.