Motif

# Motif > How good of a vibe coder are you? Real-time dashboard and assessment report from your AI coding conversations. Motif is a CLI tool that reads your Cursor and Claude Code conversations and gives you two things: a live dashboard that tracks your AI output in real-time (tokens per minute, concurrency, per-agent efficiency), and a shareable vibe report showing how you work with AI — sessions in parallel, growth over time, personality. It also discovers recurring workflows and generates personalized CLAUDE.md / .cursorrules / skills files tuned to how you actually work. ## Key Properties - No API key needed — your IDE's agent does the analysis - No server — everything stays on your machine - Open source (Apache 2.0) - Python 3.10+, installable via pip ## How It Works 1. `motif extract` gathers conversations from Cursor IDE (state.vscdb) and/or Claude Code (~/.claude/projects/) 2. `motif live` launches a real-time TUI dashboard tracking AIPM, concurrency, and per-agent efficiency as you code (currently Claude Code; Cursor coming soon) 3. `motif vibe-report` generates a self-contained HTML assessment — agent concurrency, autonomy ratio, growth scorecard, personality/catchphrases, project constellation 4. `motif analyze --prepare` runs a data pipeline (truncation, relevance filtering, token budgeting) and produces a structured payload with embedded analysis instructions. Includes the user's existing CLAUDE.md (if found) so the analysis avoids suggesting duplicate rules 5. The user's IDE agent reads the payload and performs pattern analysis — discovering skills, rules, communication style, and improvement areas 6. `motif rules` generates a reference CLAUDE.md and skill files from the analysis output. Skill files are deployed to both Cursor (`.cursor/skills/`) and Claude Code (`.claude/commands/`) directories. CLAUDE.md is never written directly — the host agent reads the reference and proposes targeted edits to the user's existing file ## Commands - `motif extract cursor|claude|all` — Extract conversations - `motif live` — Real-time AI productivity dashboard (TUI). Tracks AIPM, concurrency, per-agent efficiency. Flags: `--compact`, `--summary`, `--history`, `-i <seconds>` - `motif list` — Show extracted projects with message counts - `motif vibe-report` — Generate a shareable HTML vibe report ("Spotify Wrapped" for vibe coding) from extracted conversations. Self-contained HTML with dark theme, Chart.js charts. Sections: hero stats, agent concurrency (sweep-line algorithm), autonomy ratio timeline, project constellation, growth scorecard, personality/catchphrases. Flags: `--name`, `--output`, `--analysis` - `motif analyze --prepare` — Prepare data for agent-driven analysis. Flags: `--mode` (`full` for Personalize AI, `vibe-report` for qualitative vibe report — strips system noise, instructions-first layout), `--budget`, `--project`, `--no-filter`, `--preview`, `--stats` - `motif rules <analysis.json>` — Generate config files from analysis. Motif never writes to the user's CLAUDE.md — it generates a reference at `~/.motif/generated/CLAUDE.md` for the host agent to read and merge. `--apply` deploys skill files only. Flags: `--dry-run`, `--apply`, `-p` - `motif report <analysis.json>` — Generate summary report - `motif setup` — Install skill for agent integration (Cursor skill + Claude Code `/motif` command) - `motif update` — Check for newer versions and upgrade ## Install ``` pip install motif-cli ``` ## VS Code Extension The Motif VS Code extension brings the live dashboard into Cursor's status bar and sidebar. Color-coded AIPM indicator, session stats panel, and personal bests — no terminal required. Install from the VS Code Marketplace. ## Links - PyPI: https://pypi.org/project/motif-cli/ - GitHub: https://github.com/Bulugulu/motif-cli - VS Code Extension (Cursor/Open VSX): https://open-vsx.org/extension/motif/motif - VS Code Extension (VS Code Marketplace): https://marketplace.visualstudio.com/items?itemName=motif.motif