Oh My Agentic ScoreMeasure Your Agentic Coding Performance

Analyze Claude Code sessions across 4 dimensions — parallelism, autonomy, density, and trust — and track your journey from Base threads to fully autonomous Z-threads.

Get Started

View on GitHub

📊

4-Dimension Scoring

Evaluate sessions on More (parallelism), Longer (autonomy), Thicker (density), and Fewer (trust) — each scored 0-10 with log-normalized scaling.

🧵

7 Thread Types

Classify every session from Base conversations to Z-threads (zero-touch autonomous). Track your evolution through P, L, F, B, and C threads.

⚖️

Fair Comparison

Minimum thresholds, weighted scoring, and consistency metrics ensure short test sessions don't skew your aggregate scores.

🚀

Interactive Dashboard

Next.js dashboard with radar charts, trend lines, per-project breakdowns, and session detail views — all powered by local data.