kid0317/crewai_mas_demo/self-retrospective

Claude Code and agent skill by kid0317/crewai_mas_demo

Trust tier 2
  • community
LLM review pending
Source check pending

Checking that the SKILL.md file is still in the repository.

自我复盘思考框架。当你收到 type=retro_trigger 的邮件、或被要求"反思本周工作"/"自我复盘"时加载此 Skill。用 log_query CLI 查数据,用 LLM 推理做分析,产出结构化复盘报告和改进提案。

Install

npx vskill@latest install kid0317/crewai_mas_demo/self-retrospective

SKILL.md

自我复盘

你的任务

分析过去 7 天的工作记录,找出系统性问题,产出改进方案。

可用工具

通过 bash 调用 tools/log_query.py,支持以下查询(按需使用,不限顺序):

bash
# 整体统计
python3 tools/log_query.py stats --agent-id {agent_id} --days 7
# 任务列表(可排序)
python3 tools/log_query.py tasks --agent-id {agent_id} --days 7 --sort quality_asc --limit 5
# 某任务的 ReAct 步骤回放
python3 tools/log_query.py steps --task-id <TASK_ID> --sessions-dir workspace/{agent_id}/sessions [--only-failed]
# 人类纠正记录
python3 tools/log_query.py l1 --days 7 [--keyword "关键词"]

思考框架(非强制顺序,按需使用)

你的分析目标是回答五个递进问题:

  1. 哪些任务做得差? — 从统计和任务列表中发现模式
  2. 差在哪一步? — 下钻具体任务的执行步骤
  3. 人类怎么看? — L1 纠正记录提供独立视角
  4. 根因是什么类型? — 必须从枚举中选择(见下方)
  5. 应该改哪个文件的哪一段? — 精确到可执行的 before/after 改动

root_cause 枚举(必选其一,禁止自由文本)

枚举值含义对应改动对象
sop_gap流程/SOP 缺步骤agent.md 或 skills/*.md
prompt_ambiguity提示词/soul 指令模糊soul.md
ability_gap知识/经验不足memory.md
integration_issue与其他 Agent 协作问题agent.md 协作部分

输出格式(必须严格遵守)

产出一份 JSON,写入 workspace/shared/proposals/{agent_id}_retro_{date}.json:

json
{
"retrospective_report": {
"agent_id": "{agent_id}",
"period": "2026-04-07 ~ 2026-04-13",
"summary": "一句话总结本周主要问题",
"findings": [
{
"pattern": "描述发现的模式",
"evidence_task_ids": ["t001", "t003", "t006"],
"l1_corroboration": "人类纠正记录是否验证了这个发现"
}
]
},
"improvement_proposals": [
{
"root_cause": "sop_gap",
"target_file": "skills/product_design/SKILL.md",
"current_behavior": "当前行为描述",
"proposed_change": "具体改动描述(自然语言)",
"before_text": "定位锚点:要改的那段原文",
"after_text": "改后的完整文本",
"expected_improvement": "预期指标变化",
"evidence": ["t001", "t003"]
}
]
}

约束

  • evidence 不允许为空 — 每条提案必须有日志 ID 支撑
  • 一次复盘最多 3 条提案 — 聚焦最重要的改进
  • target_file 只允许改自己的文件:agent.md / soul.md / memory.md / skills/*.md
  • 不要自己执行改动 — 你只产出方案,由 Manager 审批后你再执行
  • 样本不足时跳过 — task_count < 5 直接报告"样本不足,跳过复盘"
  • 禁止复述式反思 — "下次要注意 XX"不是根因分析,必须指向具体文件改动

完成后

发邮件给 Manager:

  • type: retro_report
  • subject: "自我复盘完成,{N}条提案待审阅"
  • content: proposals 文件路径

⚠️ 不直接发 Human。所有对外通信必须经过 Manager。

Security

  • Security scanPassed, 100 of 100Pattern checks on the skill's files.
  • Model reviewNot runA model review is scheduled for the next rescan.

Quality evaluation

  • Task evaluationNot runNo evaluation yet. Run one in Skill Studio.

Source

Popularity

Repository stars
162
Repository forks
85
7-day trend score
25

Stars and forks are for the whole GitHub repository, not this skill alone. Refreshed 17 May 2026, 5 months ago.

Works with

Any agent vskill installs to: 56 targets in all. These 39 have a page on this site.

  • Claude Code
  • Cursor
  • GitHub Copilot
  • Windsurf
  • Codex
  • Gemini CLI
  • Cline
  • Amp
  • Roo Code
  • Goose
  • Continue
  • Kilo Code
  • Junie
  • Trae
  • OpenHands
  • Qwen Code
  • AI Agent
  • Tabnine Chat
  • Sourcegraph Cody
  • Amazon Q
  • Firebase Genkit
  • Sweep
  • Devin
  • Phind
  • Mentat
  • Claude Desktop
  • Aider
  • GPT Pilot
  • Plandex
  • bolt.new
  • v0
  • Replit Agent
  • Anse
  • TypingMind
  • BetterChatGPT
  • LibreChat
  • Warp AI
  • Pieces
  • Fig AI
Embed badge
vskill VERIFIED badge
[![vskill VERIFIED](https://verified-skill.com/api/v1/skills/kid0317/crewai_mas_demo/self-retrospective/badge)](https://verified-skill.com/skills/kid0317/crewai_mas_demo/self-retrospective)

Something wrong with this skill or its page? Tell us.