AI Engineering Fluency

(Previously known as the "GitHub Copilot Token Tracker") !AI Engineering Fluency

Track your GitHub Copilot token usage and AI Fluency across VS Code, Visual Studio, the GitHub Copilot app, and the command line. All data is read from local session logs — nothing leaves your machine unless you opt in to cloud sync, or ask an agent in the GitHub Copilot app about your stats (the numbers it reads are then sent to the model like any other tool result; see what the agent sees).

# Install from the VS Code Marketplace
ext install RobBos.ai-engineering-fluency

# VSCodium
codium --install-extension RobBos.ai-engineering-fluency
Code OSS
code --install-extension RobBos.ai-engineering-fluency
Windsurf
windsurf --install-extension RobBos.ai-engineering-fluency
Cursor
cursor --install-extension RobBos.ai-engineering-fluency
Trae (ByteDance)
trae --install-extension RobBos.ai-engineering-fluency
Kiro (AWS)
kiro --install-extension RobBos.ai-engineering-fluency
Void
void --install-extension RobBos.ai-engineering-fluency

📖 Full VS Code extension documentation

Previously known as copilot-token-tracker (deprecated): — If you still have it installed, the extension will prompt you to migrate.

📖 JetBrains plugin docs · Debugging guide

🏗️ Visual Studio Extension

Token usage tracking inside Visual Studio 2022+, reading Copilot Chat session files directly.

📖 Full Visual Studio extension documentation

⌨️ CLI

Run anywhere with Node.js — no editor required. Get usage stats, fluency scores, and environmental impact from the terminal.

npx @rajbos/ai-engineering-fluency stats

📖 Full CLI documentation

🧩 GitHub Copilot app canvas

A side-panel canvas for the GitHub Copilot app with your token usage, cost, sessions, usage charts and fluency score — styled like the VS Code extension and powered by the CLI over your local session logs. Ships as the ai-fluency-canvas Copilot plugin; click both buttons in order (each opens a form you confirm in the app).

# Or with the Copilot CLI (the app shares the same ~/.copilot config)
copilot plugin marketplace add rajbos/ai-engineering-fluency
copilot plugin install ai-fluency-canvas@ai-engineering-fluency

📖 Copilot app canvas documentation

🐚 Oh-My-Posh Segment

Display today's and 30-day token usage directly in your terminal prompt, powered by the CLI npm package.

󱊤 1.2K today · 45.3K 30d

# Install the CLI globally first
npm install -g @rajbos/ai-engineering-fluency

Then add the segment to your oh-my-posh theme:

{
  "type": "text",
  "style": "diamond",
  "leading_diamond": "\ue0b6",
  "trailing_diamond": "\ue0b4",
  "foreground": "#ffffff",
  "background": "#005ca5",
  "cache": { "duration": "5m", "strategy": "session" },
  "template": " \uec1e {{ cmd \"ai-engineering-fluency\" \"segment\" }} "
}

📖 Full Oh-My-Posh segment setup & options

💬 GitHub Copilot CLI Statusline (Experimental)

Show token usage in the GitHub Copilot CLI status bar via the experimental STATUS_LINE feature. Combines context tokens, session duration, line changes, and daily/30-day totals in one compact line.

main +2/-1 > ctx 123.5k/200.0k > ######.... > 00:12:34 > +42/-8 > 12.9M today · 1443.5M 30d

Configure %USERPROFILE%\.copilot\settings.json to point at the included scripts:

{
  "statusLine": {
    "type": "command",
    "command": "C:\\Users\\YOURUSER\\.copilot\\statusline.cmd",
    "padding": 1
  },
  "feature_flags": { "enabled": ["STATUS_LINE"] },
  "experimental": true
}

📖 Full Copilot CLI statusline setup

🔗 Self-Hosted Sharing Server

Share usage data with your team without an Azure account. Run a lightweight API server on your own infrastructure — team members configure a single endpoint URL and upload automatically via their existing GitHub session.

• Zero Azure required — SQLite + Docker, runs anywhere
• Auth via GitHub — reuses the session already held by VS Code/Copilot, no API keys
• Optional org gating — restrict uploads to GitHub org members
• Web dashboard — see aggregated usage across your team

# docker-compose.yml
services:
  sharing-server:
    image: ghcr.io/rajbos/copilot-sharing-server:latest
    ports:
      - "3000:3000"
    environment:
      - GITHUB_CLIENT_ID=...
      - GITHUB_CLIENT_SECRET=...
      - SESSION_SECRET=${SESSION_SECRET:?Missing SESSION_SECRET}
      - BASE_URL=https://copilot.example.com
    volumes:
      - sharing_data:/data
volumes:
  sharing_data:

// VS Code settings — the only thing team members configure
{
  "aiEngineeringFluency.backend.sharingServer.enabled": true,
  "aiEngineeringFluency.backend.sharingServer.endpointUrl": "https://copilot.example.com"
}

📖 Setup & configuration guide · Server developer docs

Contributing

Interested in contributing? Check out our Contributing Guide for:

• 🐳 DevContainer Setup — Isolated development environment
• 🔧 Build & Debug Instructions — How to run and test locally
• 📋 Code Guidelines — Project structure and development principles
• 🚀 Release Process — CI/CD pipelines and automated releases