Code Insights

Code Insights

  Turn your AI coding sessions into knowledge.
  Extract decisions, learnings, and prompt quality scores. Detect patterns. Get better at working with AI.

Analyzes your sessions from Claude Code, Cursor, Codex CLI, Copilot CLI, and VS Code Copilot Chat to extract structured insights — decisions with trade-offs, learnings with root causes, prompt quality with actionable feedback, and cross-session patterns that surface what's working and what's not. All stored locally in SQLite, browsable through terminal analytics and a built-in dashboard.

No accounts. No cloud. No data leaves your machine.

Claude Code users: zero-config analysis, zero cost.
Install the hook once. Every session gets analyzed automatically using your Claude subscription.

Works with Ollama — free, local, zero API keys.
If you have Ollama installed, code-insights will detect it automatically and use it for AI analysis. No account, no cost, no data leaves your machine.

What You Get

Decisions, Learnings & Prompt Quality

Each session is analyzed to extract structured insights — decisions with trade-offs and alternatives, learnings with root causes, and prompt quality scores across 5 dimensions with actionable before/after takeaways.

Cross-Session Patterns

Weekly synthesis detects friction points, effective patterns, and prompt quality trends across all your sessions. Navigate week-by-week to see how your habits evolve — and export generated rules for your CLAUDE.md or .cursorrules.

AI Fluency Score

All of the above rolls up into your AI Fluency Score — a shareable snapshot of your coding fingerprint, working style, and top patterns.

Analytics & Cost Tracking

Activity charts, cost breakdown by project and model, session types, and multi-tool usage — all in one dashboard.

Terminal…