AgentSkillsCN

retrospective

会话回顾:回顾已完成的工作,识别那些值得提炼为技能的可复用模式。在会话结束时,通过总结学习收获与潜在技能,为后续发展奠定基础。触发条件包括:“回顾”“回顾会话”“回顾”“会话回顾”“有没有可以转化为技能的东西?”“今天的回顾”。

SKILL.md
--- frontmatter
name: retrospective
description: >
  Session retrospective: review what was done, identify reusable patterns worth turning into skills.
  Use at the end of a session to extract learnings and skill candidates.
  Triggers: "振り返り", "retrospective", "retro", "セッション振り返り",
  "スキル化できるものある?", "今日の振り返り".

Retrospective Skill

Analyze the current session to extract reusable patterns, recurring workflows, and skill candidates.

Workflow

1. Session Summary

Review the conversation history and summarize:

  • What was done: features implemented, bugs fixed, tasks completed
  • What tools/commands were used repeatedly
  • What friction occurred: errors, retries, misunderstandings, manual steps

2. Pattern Detection

Look for these signals that indicate a workflow should become a skill:

Repetition signals:

  • Same sequence of commands executed 2+ times in this session
  • A workflow that was also done in previous sessions (check MEMORY.md)
  • Manual steps that could be automated (e.g., "start server, then check proxy, then curl endpoint")

Friction signals:

  • Steps where Claude had to retry or the user had to correct
  • Configuration that was forgotten and had to be redone
  • Verification steps that were skipped and caused issues later

Complexity signals:

  • Multi-step workflows with a specific order (build → test → deploy)
  • Domain-specific knowledge needed (project conventions, API patterns)
  • Checklists that need to be followed consistently

3. Evaluate Candidates

For each candidate pattern, assess:

CriteriaQuestion
FrequencyWill this be used again? (weekly+)
ComplexityIs it more than 2 trivial steps?
Error-proneAre steps likely to be forgotten or done wrong?
GenericDoes it apply across projects or just this one?

Score each candidate:

  • 3+ criteria met → Strong skill candidate
  • 2 criteria met → Consider as skill
  • 1 or fewer → Not worth a skill (just document in MEMORY.md)

4. Report

Present findings in this format:

markdown
## Session Retrospective

### What was done
- [summary of work]

### Skill candidates found

#### 1. [Pattern name] ⭐ Strong candidate
- **What:** [description of the workflow]
- **Why:** [which signals triggered: repetition/friction/complexity]
- **Frequency:** [how often this would be used]
- **Example from this session:** [concrete example]

#### 2. [Pattern name] 🔶 Consider
- **What:** [description]
- **Why:** [signals]
- ...

### Other learnings (for MEMORY.md)
- [insights that aren't skill-worthy but worth remembering]
- [project-specific conventions discovered]
- [gotchas encountered]

5. Create Skills (with permission)

For strong candidates, ask the user:

code
Found N skill candidates. Create them?
1. ⭐ [name] — [one-line description]
2. ⭐ [name] — [one-line description]
3. 🔶 [name] — [one-line description]

If approved, create each skill at ~/.claude/skills/<name>/SKILL.md with:

  • YAML frontmatter (name, description with triggers)
  • Clear workflow steps
  • Important rules section

6. Update Memory

Append session learnings to the project's MEMORY.md:

  • New patterns discovered
  • Gotchas encountered
  • Project conventions clarified

Skill Creation Checklist

When creating a new skill from a detected pattern:

  • File: ~/.claude/skills/<name>/SKILL.md (uppercase SKILL.md)
  • YAML frontmatter with name, description, and Japanese + English triggers
  • Clear step-by-step workflow
  • "Important Rules" section with do's and don'ts
  • Concise — under 150 lines if possible

Important Rules

  • DO review the full session, not just the last few messages
  • DO check existing skills (~/.claude/skills/) to avoid duplicates
  • DO check MEMORY.md for patterns that appeared in previous sessions too
  • DO NOT create skills for one-off tasks — only recurring patterns
  • DO NOT create skills without user approval
  • Present candidates clearly so the user can decide which to create