메뉴
Runs 3 isolated reviewer sub-agents; consolidates PASS/BLOCK verdict by majority. Use when user requests independent review of code changes, pull requests, design documents, or release notes.
Codex 또는 Claude로 설치 이 Prompt를 복사해 Codex, Claude 또는 다른 어시스턴트에 붙여 넣으면 Skill 페이지를 검토하고 설치를 진행할 수 있습니다.
SOC 직업 분류 기준
Scaffolds production-ready monorepo projects using the backbone CLI: configures workspace packages, wires build tooling, sets up CI pipelines, and initializes framework/backend/CMS integrations. Use when creating, bootstrapping, or initializing a new application, project, starter template, or monorepo from scratch. Trigger terms: scaffold, bootstrap, init, new repo, new app, boilerplate, starter, greenfield.
Resolves task dependencies, generates machine-actionable delegation specs, structures phased subtask plans for multi-agent work. Use when writing delegation specs, resolving task dependencies, building phased subtask plans for multi-agent work, assigning work to sub-agents, or partitioning a feature into parallelizable phases.
Reviews mature LESSONS-LEARNED.md entries, rewrites them as permanent rules in skill/instruction files, archives graduated lessons. Use when graduating lessons into skills, promoting validated lessons, updating skills from past learnings, archiving mature lessons, codifying repeated patterns, or cleaning up a crowded LESSONS-LEARNED.md.
Logs sessions, tracks activity, records delegation decisions, and stores review/dispute outcomes as NDJSON audit trails. Use when logging session activity, tracking work, recording decisions, building audit trails, capturing delegation history, or running pre-response verification checklists. Trigger terms: log, track activity, audit trail, session record, delegation log, NDJSON
Enforces React-specific patterns: functional components with hooks, TypeScript prop interfaces, CSS Modules co-location, React Testing Library behavioral tests. Use when creating React components, writing custom hooks, structuring component folders, applying RTL test patterns, or wiring TypeScript prop types. Trigger terms: React, .tsx, component, hook, RTL, jsx, useState, useEffect, prop interface
Implements technical SEO: meta tags, JSON-LD structured data, sitemaps, crawlability fixes. Use when adding schema markup, JSON-LD, robots.txt updates, canonical URLs, Open Graph tags, or improving crawlability.
| name | panel-majority-vote |
| description | Runs 3 isolated reviewer sub-agents; consolidates PASS/BLOCK verdict by majority. Use when user requests independent review of code changes, pull requests, design documents, or release notes. |
Inputs: <runRoot>, <panelKey> (filesystem-safe), question text, artifact list. Panel dir default: <runRoot>/panel/.
| File | Path |
|---|---|
| Prompt payload (optional) | <panelDir>/<panelKey>-panel-prompt.md |
| Raw reviewer outputs | <panelDir>/<panelKey>-reviewer-outputs.md |
| Consolidated report | <panelDir>/<panelKey>.md |
Validate scope — every artifact path under <runRoot>; list sufficient to answer question.
Spawn 3 reviewers in parallel — start three isolated subagents with identical prompts. Spawn 3 reviewers using runSubagent with identical prompts; each reviewer receives the same question, artifact list, and constraints but runs in isolation. Required reviewer output sections (no others): VERDICT: PASS | BLOCK, MUST-FIX:, SHOULD-FIX:, QUESTIONS:, TEST IDEAS:, CONFIDENCE: low | med | high.
Persist outputs — write <panelDir>/<panelKey>-reviewer-outputs.md with a header (run root, panel key, question, artifacts) and each reviewer output verbatim, separated.
Consolidate — parse each reviewer output for its VERDICT: line and count PASS votes. Overall verdict = PASS if pass_count ≥ 2; otherwise BLOCK. Deduplicate MUST-FIX: and SHOULD-FIX: items and annotate each item with (N/3 reviewers).
# count PASS/BLOCK from combined outputs
pass_count=$(grep -o "VERDICT: PASS" panel/run123-reviewer-outputs.md | wc -l)
block_count=$(grep -o "VERDICT: BLOCK" panel/run123-reviewer-outputs.md | wc -l)
verdict=$([ "$pass_count" -ge 2 ] && echo PASS || echo BLOCK)
# emit a minimal JSON summary using jq (install jq if needed)
jq -n --arg panel_key "run123-panel" --arg verdict "$verdict" --argjson pass_count $pass_count --argjson block_count $block_count '{panel_key:$panel_key, verdict:$verdict, pass_count:$pass_count, block_count:$block_count}' > panel/run123-summary.json
<panelDir>/<panelKey>.md with the minimal structure below and reference the generated summary.<panelKey> — Verdict: PASS | BLOCK (pass_count/block_count)MUST-FIX and SHOULD-FIX itemspanel/run123-summary.jsonPrint summary — overall verdict + vote tally + report path.
Log (⛔ hard gate) — call the observability-logging skill with panel_key, verdict, pass_count, block_count, must_fix, should_fix, reviewer_model, weighted, attempt, tracker_issue, artifacts_count, report_path for verification.
| Resource | Purpose |
|---|---|
panel-report.template.md | Report template for step 5 |
REFERENCE.md | Weighted consensus variant and weighting details |
| observability-logging skill | Panel logging command (step 7) |
| team-lead-reference skill | Model routing and dispute protocol |