菜单
Provide structured feedback with severity classification
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
基于 SOC 职业分类
Close a PROJECT_PLAN with verifiable-assertion gate (V-tests + commits + retrospective), deferred-surface scan for next-plan handoff, and status transition (ACTIVE → COMPLETE / CLOSED / ABANDONED / SUPERSEDED). Strict counterpart to /aget-create-project (D71 Layer 2). Closes asymmetric verb-pair gap at PROJECT_PLAN lifecycle.
Commit a selected candidate goal into a durable, governed Goal artifact. The D71-Strict singular committer of the propose-goals→create-goal verb pair (L1067/L1085). Two-tier (REQ-3): committed goals = a structured section in governance/GOALS.md; aspirational = agent-internal .aget/goals/aspirational.jsonl. Direct authoring of a committed-Goal section is PROHIBITED once this skill is invoked. Implements AGET_GOAL_SPEC v0.2.0 (CAP-GOAL-001..011).
Scaffold approved initiative manifests at planning/initiatives/INIT-*.md. Mirrors /aget-create-project STRICT (D71 Layer 2) — direct Write/Edit to planning/initiatives/ is PROHIBITED once this skill is invoked. Implements SOP_initiative.md (graduated procedure) + future AGET_INITIATIVE_SPEC. Consumes PROPOSAL_init_*.md substrate produced by /aget-propose-initiative.
Create lightweight initiative proposals (PROPOSAL_init_*.md) before full INIT-*.md scaffolding. Follows 'propose -> review -> approve -> create' governance pattern. Advisory enforcement. Implements AGET_INITIATIVE_SPEC v1.0.1 CAP-INIT-PROP-001..012.
Record principal authorization (GO) with Healthy Friction enforcement of the principle-triad (spec+verify-first, coherence-next, evidence-driven); release agent execution within authorized scope. Authorization-only — does not execute the authorized work.
Generate standalone narrative documents from session artifacts, optimized for a target consumption medium (NotebookLM, slides, memo, share-draft). Use when the user says "create briefing", "create a briefing for NotebookLM", "make this listenable", "summarize session for [medium]", "prepare a listen", "brief me on [topic]", or "create a summary I can share".
| name | aget-provide-feedback |
| description | Provide structured feedback with severity classification |
| archetype | reviewer |
| allowed-tools | ["Read","Glob"] |
Provide structured feedback with severity classification. Balances constructive criticism with positive observations.
When this skill is invoked:
Review Work Product
Classify Feedback
Make Actionable
Balance
## Feedback: [Work Product]
### Summary
| Severity | Count |
|----------|-------|
| Blocking | [N] |
| Suggestion | [N] |
| Nitpick | [N] |
| Praise | [N] |
---
### Blocking Issues
1. **[Location]**
- Issue: [Description]
- Impact: [Why this blocks]
- Suggestion: [How to resolve]
---
### Suggestions
1. **[Location]**
- Observation: [What could be better]
- Rationale: [Why it matters]
- Suggestion: [Specific improvement]
---
### Nitpicks
1. **[Location]**: [Minor observation]
---
### Positive Observations
1. **[Location]**: [What was done well]
2. **General**: [Overall strength]
---
### Overall Assessment
[Brief summary of quality level and key points]
| Severity | Criteria | Examples |
|---|---|---|
| Blocking | Prevents functionality, correctness, or compliance | Bug, security issue, spec violation |
| Suggestion | Improves quality, maintainability, clarity | Refactoring opportunity, better naming |
| Nitpick | Style, preference, minor | Formatting, comment wording |
| Praise | Notable positive | Elegant solution, thorough testing |