.claude/commands/session-insights.md
code:md
---
description: Extract key project insights and update CLAUDE.md with actionable improvements
---
Extract project-specific insights from the session to update CLAUDE.md with ULTRATHINK.
**3-Axis Analysis**
- **Goal Achievement**: Objective completion effectiveness
- **Efficiency**: Process and resource optimization
- **User Satisfaction**: Emotional experience and frustration analysis
**Output Requirements**
**ALWAYS** use user language for output.
**Analysis Ratings**
Rating guide: ★☆☆☆☆ Failed/Very poor | ★★☆☆☆ Poor | ★★★☆☆ Adequate | ★★★★☆ Good | ★★★★★ Excellent
**Steps**
0. **Preparation**: Review session context and notes
- Understand project goals and user expectations
- Identify key events and actions taken during the session
**Output Template:**
`
Starting session analysis...
`
1. **Analysis: Goal Achievement**: Evaluate objective completion
- Task completion status and quality
- Project outcomes delivered
- Goals missed and blockers
**Output Template:**
`
## Goal Achievement: ★★★☆☆ (3/5)
- Reason 1
- Reason 2
- Reason 3
`
2. **Analysis: Efficiency**: Assess process optimization
- Effective vs ineffective approaches
- Resource waste patterns
- Future optimization opportunities
**Output Template:**
`
## Efficiency: ★★★☆☆ (3/5)
- Reason 1
- Reason 2
- Reason 3
`
3. **Analysis: User Satisfaction**: Analyze emotional experience
- AI actions triggering negative reactions
- Frustration patterns and root causes
- Expectation violations and communication issues
**Output Template:**
`
## User Satisfaction: ★★★☆☆ (3/5)
- Reason 1
- Reason 2
- Reason 3
`
4. **Thinking: Synthesize Insights**: Two improvement types:
**Codebase**: Lint/test/refactor changes enhancing AI coding
**LLM Context**: CLAUDE.md updates improving AI understanding
5. **Report: Present Insights for LLM Context Improvements**: Display only the most critical improvements for user approval:
**Output Template:**
`
## LLM Context Improvements
- ADD: New context to add: How it improves AI understanding
- DELETE: Context to remove: Why it causes problems
- MODIFY: Context to modify: Current issue → Improved version
Do you approve these LLM context improvements for CLAUDE.md updates?
`
Principles: Extracted insights only, project-relevant, actionable, organized by type
6. **Action: Review CLAUDE.md**: Check structure and style
7. **Action: Apply LLM Improvements**: Update CLAUDE.md with approved insights
8. **Report: Present Codebase Improvements**:
`
## Codebase Improvements Suggestions
- Specific codebase change: How it enhances AI coding quality
- Example: Add ESLint rules for consistent imports - Reduces AI confusion about module resolution
Would you like me to create GitHub issues for these codebase improvements?
`
9. **Action: Create Issues (If Approved)**: Make GitHub issues with detailed descriptions:
`
Title: Improvement Type: Brief description
Example: "Code Quality: Add ESLint rules for consistent imports"
Body:
## Context
Identified during AI coding session retrospective
## Problem
Root cause that led to AI coding issues
## Proposed Solution
Specific change and implementation approach
## Expected Benefit
How this enhances future AI coding quality
## Priority
Low/Medium/High based on impact
`
10. **Finalize**: Ensure consistency
**Guidelines**
- Project-specific insights only
- Root causes, not symptoms
- Concrete improvements, not vague aspirations
- Prevent frustration patterns
- Verify CLAUDE.md consistency