Our review
Analyzes Claude Code conversation history to identify friction patterns and suggest improvements to configuration.
Strengths
- Leverages real conversations for context-aware suggestions
- Generates actionable improvement proposals in a file
- Automates detection of recurring issues
- Recommends updates to CLAUDE.md and skills
Limitations
- Requires access to conversation history
- Can be slow with large volumes of data
- Does not automatically apply changes
Use this skill when you or your project faces repeated friction with the Claude agent and you want systematic improvements.
Do not use if you have no relevant conversation history or prefer manual configuration adjustments.
Security analysis
SafeThe skill only reads local conversation history and writes a local markdown file; it uses no network calls, destructive commands, or data exfiltration. The bash commands are limited to listing files, and subagents only analyze content locally.
No concerns found
Examples
/self-improvement/self-improvement last 3 days/self-improvement refactoringname: self-improvement description: Analyze conversation history to find friction patterns and suggest CLAUDE.md/skill improvements. Use when user wants to review what went wrong across sessions and systematically improve. (user) allowed-tools: Read, Bash, Grep, Glob, Task
Self-Improvement - Learn from History
Analyze Claude Code conversation history to find friction patterns, check what's already fixed, and suggest improvements.
Instructions
Phase 1: Find Conversations
# Get project directory (encode current path with dashes)
ls ~/.claude/projects/
# List today's conversations (or adjust -mtime for longer range)
ls -lt ~/.claude/projects/<encoded-path>/*.jsonl | head -20
Phase 2: Parallel Analysis
Spawn Task agents (subagent_type: general-purpose) to analyze conversations in parallel. Each agent reads one .jsonl file and extracts:
- What was user trying to accomplish?
- Problems/friction that occurred (include user quotes showing frustration)
- What worked well?
- Repeated patterns or inefficiencies
For large files (>500KB), prioritize those - they contain the meatiest sessions.
Phase 3: Synthesize
After all agents complete, combine findings:
- Top friction patterns ranked by frequency
- What worked well (don't lose these)
- User frustration quotes (raw evidence)
Generalize aggressively. Look for the meta-pattern behind specific issues.
Phase 4: Cross-Reference
Read current documentation:
- Project CLAUDE.md
- User ~/.claude/CLAUDE.md
- Skills in .claude/skills/ and ~/.claude/skills/
For each friction pattern, classify:
- Already fixed - note location
- Still missing - needs addition
Phase 5: Output
Create a markdown file (e.g., CLAUDE_IMPROVEMENTS.md) with:
# Suggested CLAUDE.md Improvements
Based on analysis of N conversations from [date range].
## 1. [Issue Name]
**Problem:** [Description of friction]
**Suggested addition to [location]:**
\```markdown
[Proposed text]
\```
---
## Already Fixed
| Issue | Where |
|-------|-------|
| ... | ... |
---
## Potential Skills
| Skill | Purpose |
|-------|---------|
| ... | ... |
---
## Raw Friction Log
- "user quote 1"
- "user quote 2"
Do not apply changes - create the file for user review.
Usage Examples
/self-improvement
/self-improvement last 3 days
/self-improvement refactoring
Task Prioritizer
Productivity
Prioritizes your tasks using Eisenhower, ICE, and RICE frameworks.
Weekly Status Report Generator
Productivity
Generate structured and concise weekly status reports.
Daily Standup Report
Productivity
Generates structured and concise daily standup reports.