UseAI

Track your AI coding sessions with privacy-first analytics.

UseAI is a local-first MCP server that records how you use AI coding tools -- session duration, languages, task types, and streaks -- without ever seeing your code. Think of it as Wakatime for AI coding.

Features

• Session tracking -- automatically records when you start and stop using AI tools
• Streak tracking -- daily coding streaks with global leaderboard
• Evaluation metrics -- sessions scored using the SPACE framework for prompt quality, context, independence, and scope
• Local dashboard -- built-in web UI served from the daemon (useai serve)
• Public profile & leaderboard -- opt-in shareable profile at useai.dev with global AI proficiency rankings
• Privacy-first -- everything stays in ~/.useai/ on your machine, zero network calls from the MCP server
• Ed25519 signed chain -- every session record is cryptographically signed for tamper evidence
• 30+ AI tools supported -- Claude Code, Cursor, Windsurf, VS Code, Codex, Gemini CLI, GitHub Copilot, Aider, Cline, Zed, Amazon Q, JetBrains/Junie, Goose, Roo Code, and many more

Quick Start

npx @devness/useai

This installs the MCP server and configures it for your AI tools automatically.

Manual Setup

claude mcp add useai -- npx -y @devness/useai

Add to .cursor/mcp.json:

{
  "mcpServers": {
    "useai": {
      "command": "npx",
      "args": ["-y", "@devness/useai"]
    }
  }
}

Add to your VS Code MCP settings:

{
  "mcp": {
    "servers": {
      "useai": {
        "command": "npx",
        "args": ["-y", "@devness/useai"]
      }
    }
  }
}

Add to your Windsurf MCP config:

{
  "mcpServers": {
    "useai": {
      "command": "npx",
      "args": ["-y", "@devness/useai"]
    }
  }
}

No API key needed. The MCP server runs 100% locally.

How It Works

UseAI runs as an MCP (Model Context Protocol) server. When your AI tool starts a conversation, it calls useai_start. During the session, periodic useai_heartbeat calls keep the timer alive. When the conversation ends, useai_end seals the session with a cryptographic signature.

All data is written to ~/.useai/ as JSONL files. The MCP server makes zero network calls.

| MCP Tool | What it does | |----------|--------------| | useai_start | Begin tracking a session | | useai_heartbeat | Keep-alive during long sessions | | useai_end | End session, record milestones and evaluation |

Daemon Mode

For tools that support HTTP-based MCP (StreamableHTTP), UseAI can run as a background daemon on 127.0.0.1:19200. This allows multiple AI tool sessions to connect concurrently to the same tracking instance:

useai serve                   # Start daemon + local dashboard

The setup wizard auto-configures the right mode (stdio or daemon) for each tool.

Evaluation Frameworks

UseAI uses configurable evaluation frameworks to score each AI coding session. The framework controls what rubric the AI model uses when self-assessing session quality, producing a 0-100 session score.

| Framework | Description | |-----------|-------------| | SPACE (default) | Based on the SPACE developer productivity framework by GitHub/Microsoft Research. Weighted rubrics across Communication (prompt quality, context), Efficiency (independence), and Performance (scope). | | Basic | Simple equal-weight average across all dimensions. No detailed rubric guidance. |

Session scores feed into the AI Proficiency Score (APS) -- a 0-1000 composite score across five dimensions: Output, Efficiency, Prompt Quality, Consistency, and Breadth.

Change your framework:

useai config --framework space    # recommended
useai config --framework raw      # basic/no rubric

The setup wizard (npx @devness/useai) also lets you pick a framework during installation.

What Gets Tracked

• Which AI tool you're using (Cursor, Claude Code, etc.)
• Session duration and task type (coding, debugging, testing, etc.)
• Programming languages used
• File count (number only, never file names)
• Generic milestone descriptions (privacy-filtered by design)
• Self-evaluation metrics (prompt quality, task outcome, independence)

Never tracked: your code, prompts, AI responses, file names, file paths, or directory structure.

What Gets Synced

When you run useai sync, full session records (not just aggregates) are sent to the server. This includes all fields above plus private_title and project name. See PRIVACY.md for the exact payload and what's publicly visible vs. private.

CLI

useai stats         # View local stats: streaks, hours, tools, languages
useai status        # See everything stored on your machine
useai sync          # Sync sessions to useai.dev
useai serve         # Start local analytics dashboard
useai config        # Manage settings

Install globally:

npm install -g @devness/useai-cli

Architecture

UseAI is a monorepo with the following open source packages:

packages/
  shared/       Core types, constants, Ed25519 crypto chain
  mcp/          MCP server + HTTP daemon (published as @devness/useai)
  cli/          CLI tool (published as @devness/useai-cli)
  ui/           Shared React component library
  dashboard/    Local analytics web UI
  web/          Public website (useai.dev)

Tech stack: TypeScript, pnpm workspaces, Turborepo, Vitest, React 19, Tailwind v4.

Privacy

UseAI is designed with privacy as architecture, not just policy:

• Local-first -- MCP server writes to disk, never to the network
• Open source -- audit exactly what gets recorded
• Cryptographic chain -- Ed25519 signed hash chain for tamper evidence
• Opt-in sync -- data only leaves your machine when you choose
• You own your data -- export or delete at any time

For a complete list of every field captured, what happens when you sync, and what's visible on your public profile, see PRIVACY.md. For details on the cryptographic chain, see SECURITY.md.

Contributing

We welcome contributions! See CONTRIBUTING.md for setup instructions and guidelines.

Links

• Website: useai.dev
• GitHub: devness-com/useai
• npm: @devness/useai
• Explore supported tools: useai.dev/explore

License

MIT