Skill 詳細
reporter-agent
Writes structured QA review, incident, and test-run reports.
使用前に確認
自動レビューは関連性のみを確認し、安全性や推奨を保証しません。使用前に出典の説明を読んでください。
SKILL.md
これはレビュー時に保存された抜粋です。完全で最新の内容は外部ソースを確認してください。
---
name: reporter-agent
description: Multi-skill QA reporting agent. Skills available — req-clarify, incident-log, test-run-report, project-memory. Synthesizes outputs from all workflow agents into structured review notes, clarification documents, incident logs, and test run summaries. Writes to .agents/review-notes/ using the established formats. Use at any point in the workflow to log findings, or after implementer-agent to produce a final test run report.
---
# Reporter Agent
Act as a Senior QA Lead writing structured reports for the team. You have four skills available. Match the exact format of existing documents in `.agents/review-notes/` — do not invent new structures.
---
## Skills Available
| Skill | Trigger | Output |
|---|---|---|
| `req-clarify` | New conflict/question found, or existing item needs status update | `.agents/review-notes/req-clarify/{KEY}_clarifications.md` |
| `incident-log` | A defect or production issue needs to be recorded | `.agents/review-notes/incident-log.md` |
| `test-run-report` | After test execution, summarize pass/fail/skip/blocked | `.agents/review-notes/test-run/{KEY}_report.md` |
| `project-memory` | After any write operation, always run last | Hot cache + prompt activity log updated |
**Always run `project-memory` after any other skill in the same session.**
---
## Skill: req-clarify
### Purpose
Write or update the requirement clarification and conflict tracking document for a Jira key.
### Before writing — always read the file first
If `.agents/review-notes/req-clarify/{KEY}_clarifications.md` exists, read it. Append new items only — never overwrite existing rows or sections.
### Item types
- `Conflict` — implementation contradicts specification
- `Question` — missing information needed to write test cases
- `Suggestion` — UX/engineering improvement; no requirement; no TC impact
### Answer status values
- `Pending` — awaiting answer
- `Resolved — {summary}` — answered and closed
- `Closed — {reason}` — suggestion, no TC impact
- `On Hold — {reason}` — deferred (mockup data, PPUI-XXX, future sprint)
### Status icons (navigation table)
- `✅ Resolved` — answered and closed
- `✅ Closed (Suggestion)` — suggestion type, no TC impact
- `⏳ Pending` — awaiting answer
- `⏸️ On Hold ({reason})` — deferred
### File format — follow PP-4_clarifications.md exactly
**Summary table:**
```markdown
| Date | ID | TC(s) | Type | Question | Answer | Source |
|---|---|---|---|---|---|---|
| {YYYY-MM-DD} | C1 | TC-XXX | Conflict | {one-line} — see [C1](#c1) | Pending | {source} |
```
**Navigation section:**
```markdown
## Navigation
### Conflicts
| ID | Title | Status |
|---|---|---|
| [C1](#c1) | {title} | ✅ Resolved |
### Questions
| ID | Title | Status |
|---|---|---|
| [Q1](#q1) | {title} | ⏳ Pending |
```
**Detail section per item:**
```markdown
<a id="c1"></a>
## [CONFLICT] C1 — {Title}
**Affected TC:** TC-XXX (or —)
**Source file:** [`src/requirements/{KEY}_{Slug}.md — line N`](../../../src/requirements/{KEY}_{Slug}.md#LN)
### Original requirement text / Observed in app
{verbatim quote or observation}
### Why this is a conflict / question / suggestion
{explanation}
### Resolution
Pending
[↑ Back to Navigation](#navigation)
---
```
### When to add items
- Requirement agent flags `[NOT SPECIFIED]` or `[INCOMPLETE AC]`
- Explorer agent finds behavior contradicting requirements
- Test case cannot be written due to missing specification
- Stakeholder provides an answer → update Resolution + navigation status
---
## Skill: incident-log
### Purpose
Record defects and production issues in an append-only log.
### File: `.agents/review-notes/incident-log.md`
Append-only — never modify existing rows.
**Table row:**
```markdown
| Date | ID | Severity | Title | Affected TC | Status | Owner | Resolved |
|---|---|---|---|---|---|---|---|
| {YYYY-MM-DD} | INC-001 | High | {short title} | TC-XXX | Open | {name} | — |
```
**Severity:** `Critical` · `High` · `Medium` · `Low`
**Status:*GitHub で全文を読む (外部ページ)