Setting the file. One moment.
README · Retrospecting · bitwarden/ai-plugins · Skills Docs
Comprehensive analysis of Claude Code sessions to identify successful patterns, problematic areas, and opportunities for improvement.
Back to the top of the page Running Work Transitions
This skill analyzes completed Claude Code sessions by examining:
Git history (commits, diffs, file changes)
Claude Code native session logs from ~/.claude/projects/{project-dir}/{session-id}.jsonl
Code quality metrics (tests, compilation, standards)
Your direct feedback about the session
It produces a structured retrospective report with:
Quantitative metrics (files changed, test coverage, tool usage)
Qualitative insights (what worked, what didn’t, why)
Actionable recommendations for future sessions
Reusable patterns and anti-patterns
Invoke this skill when you want to:
Review what was accomplished in a session
Understand what went well and what could improve
Get feedback on workflow effectiveness
Document lessons learned for future reference
Analyze a particularly successful or challenging session
“Can you do a retrospective on what we just accomplished?”
“How did that session go?”
“Analyze the last 2 hours of work”
“What could we improve about how we worked together?”
Scope Definition - You’ll be asked:
Time range or commit range for the session
What you were trying to accomplish
How detailed of an analysis you need (quick/standard/comprehensive)
Data Collection - The skill will:
Analyze git history for the session timeframe
Parse Claude Code native session logs from ~/.claude/projects/{project-dir}/{session-id}.jsonl
Examine changed files and code quality
Gather your feedback through targeted questions
Analysis & Report - You’ll receive:
Structured retrospective report (markdown format)
Evidence-based findings with specific examples
Prioritized recommendations
Patterns worth replicating or avoiding
Validation & Refinement - You can:
Confirm the analysis matches your experience
Add pain points that were missed
Prioritize which improvements matter most
Request configuration updates based on findings
Cleanup (Optional) - You can:
Delete the session log files after analysis
Keep them for future reference
Quick retrospective : 5-10 minutes (brief summary)Standard retrospective : 15-20 minutes (balanced analysis)Comprehensive retrospective : 30+ minutes (detailed deep-dive) The skill will recommend a depth based on your session size (commits, log volume, complexity).
Quick Retrospective - Concise summary with:
Highlights (top successes)
Challenges (main issues)
Key learnings
2-3 action items
Comprehensive Retrospective - Detailed analysis with:
Executive summary with metrics
Success patterns with evidence
Pain points with root causes
Workflow optimization analysis
Prioritized recommendations
Patterns for future reference
All reports are saved to:
The session ID links the report to the original conversation logs for traceability.
You can request a specific depth level:
Quick : Fast summary for simple sessionsStandard : Balanced analysis for typical sessionsComprehensive : Deep dive for complex or significant sessionsCustom : Focus on specific areas you define The skill will suggest an appropriate depth based on session metrics, but you can override.
You can request focus on specific aspects:
“Focus on code quality and testing”
“Analyze communication effectiveness”
“Look for security considerations we might have missed”
“Compare this session to previous similar work”
If the retrospective identifies patterns that should become standard practice, the skill will:
Suggest updates to .claude/CLAUDE.md, SKILL.md files, or agent definitions
Show you the proposed changes
Apply them with your approval
This creates a continuous improvement loop.
Successful patterns and anti-patterns are extracted into reusable libraries (future feature):
After generating a report, you can optionally delete the session logs to reduce repository size. The retrospective report preserves key insights, so logs are often not needed afterward.
When asked “What were you trying to accomplish?”, be specific:
Good : “Refactor authentication to use biometric providers and add unit tests”Less helpful : “Work on authentication” The skill asks for feedback about friction areas. Candid input leads to better insights:
Where did you get confused?
What took longer than expected?
What would you change about the workflow?
After major features : Capture complex implementation insightsWeekly/sprint boundaries : Track progress and improvement trendsAfter challenging sessions : Learn from difficultiesAfter smooth sessions : Identify what made them effective Retrospectives are most valuable when recommendations are acted upon:
Implement high-priority improvements in next session
Update configuration files with better practices
Share learnings with team (if applicable)
Session logs are automatically generated by Claude Code in ~/.claude/projects/{project-dir}/
Project directory is derived from your working directory (slashes replaced with dashes)
Example: /Users/user/project → ~/.claude/projects/-Users-user-project/
You can still do a retrospective using git history + your feedback if logs are unavailable
Provide more specific session goals upfront
Add detailed feedback when prompted
Request a comprehensive retrospective for deeper analysis
The validation step (Step 6) is exactly for this
Tell the skill what’s missing or incorrect
It will refine the analysis based on your input
Request a “quick” retrospective for large sessions
The skill will summarize logs instead of reading them fully
Focus on specific areas rather than comprehensive analysis
Git commits and diffs visible in your repository
Claude Code native session logs from ~/.claude/projects/{project-dir}/{session-id}.jsonl
File contents of changed files
Your explicit feedback responses
All analysis happens locally in your session
Reports stored in your repository (${CLAUDE_PROJECT_DIR}/.claude/skills/retrospecting/reports/)
No data sent to external services
Logs can be deleted after retrospective if desired
If your session involved sensitive data:
Review generated reports before committing them
Retrospective can be run without committing reports
You can request specific sections be excluded from the report
User : “Can you do a quick retrospective on what we just did?”
Skill checks: 8 commits, 3 log files, ~45 minutes of work
Suggests: Quick retrospective (5-10 min)
Analyzes git changes, skims logs, asks 2-3 questions
Generates quick report with highlights, challenges, learnings
Takes ~7 minutes total
User : “I want a detailed retrospective on the entire authentication refactor we did today”
Skill checks: 34 commits, 12 log files, ~4 hours of work
Recommends: Comprehensive retrospective (30 min)
Deep analysis of git history, detailed log parsing, code quality metrics
Extensive user feedback gathering (7-10 questions)
Generates full report with metrics, patterns, recommendations
Suggests 3 configuration improvements for .claude/CLAUDE.md
Offers to delete logs after report validated
User : “Analyze why testing took so long this session”
Skill identifies custom focus: testing workflow
Examines test-related commits and file changes
Searches logs for test failures, retry patterns
Asks targeted questions about testing pain points
Generates focused report on testing efficiency with specific recommendations
Planned improvements to this skill:
Automated CI integration (run retrospectives on PR merge)
Trend analysis across multiple sessions
Pattern library with searchable best practices
Comparative analysis (this session vs similar past sessions)
Team retrospectives (multi-user session analysis)
Found an issue or have a suggestion? Update the skill:
Modify ${CLAUDE_PROJECT_DIR}/.claude/skills/retrospecting/SKILL.md (Claude’s instructions)
Modify this README.md (user documentation)
Add context files in contexts/ for additional analysis frameworks
Update templates in templates/ for different report formats
Previous
Bundled file Session Analytics