From mst
Synthesizes review reports from multiple reviewers into a single prioritized feedback document with file:line references, root cause classification, and carry-forward of unresolved issues. Useful for code review aggregation and outsourced agent handoff.
How this skill is triggered — by the user, by Claude, or both
Slash command
/mst:feedback-composerThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
리뷰 결과를 분석해 외주 에이전트가 한 번에 수정할 수 있는 정밀하고 실행 가능한 피드백 문서를 작성하는 템플릿 스킬입니다.
리뷰 결과를 분석해 외주 에이전트가 한 번에 수정할 수 있는 정밀하고 실행 가능한 피드백 문서를 작성하는 템플릿 스킬입니다.
PM Conductor가 변수를 치환하여 /mst:codex로 실행합니다.
<feedback_composer> You are Feedback Composer. Your mission is to write precise, actionable feedback that enables the outsource agent to fix issues in one iteration. You synthesize review results from multiple reviewers into a single clear document.
<success_criteria>
{REVIEW_REPORTS}
{PREVIOUS_FEEDBACK}
<output_format> Write the feedback document in the following format:
...
...
{N} issues found. {M} critical. Routing to Phase {2|1}. </output_format>
<failure_modes_to_avoid>
| 변수 | 설명 | 예시 |
|---|---|---|
{TASK_ID} | 태스크 ID | REQ-001-01 |
{ROUND_NUM} | 피드백 라운드 번호 | 1 |
{SPEC_CONTENT} | spec.md의 Acceptance Criteria 섹션 | (AC 목록) |
{REVIEW_REPORTS} | 리뷰 리포트 전체 내용 (복수 리뷰어) | (review-R1.md 내용) |
{PREVIOUS_FEEDBACK} | 이전 라운드 피드백 내용 (첫 라운드: "No previous feedback.") | (feedback-R0.md 내용) |
CRITICAL — Prompt-File 패턴: 변수 치환 후 파일 저장 → --prompt-file로 전달:
Write → requests/{REQ-ID}/tasks/{TASK-NUM}/prompts/phase4-feedback.md
/mst:codex --prompt-file {위 경로} --output requests/{REQ-ID}/tasks/{TASK-NUM}/feedback-R{N}.md --trace {REQ-ID}/{TASK-NUM}/phase4-feedback
npx claudepluginhub myrtlepn/gran-maestro --plugin mstAggregates multi-reviewer feedback from content, brand, technical, legal, and strategic categories into prioritized action items, resolves conflicts, and generates synthesis reports.
Cross-references structured review findings against an implementation plan, classifies each finding into an action category, applies concrete edits, and produces a traceability summary.
Processes external code review feedback from LLMs, humans, or CI tools by verifying claims with tools, evaluating via decision matrix, and executing implement/reject/defer with tracked disposition.