Copilot Metrics Viewer · v3.2.0 · MIT License · Source
Self-hosted, read-only dashboard for GitHub Copilot usage metrics. Reads from the Copilot Usage Metrics API and displays charts/tables. Does NOT generate code, make automated decisions, or modify repositories.
- Organization → Members: Read
- Organization → Copilot Metrics: Read
- Organization → Copilot Seat Management: Read
No write permissions requested. Optional AI chat needs Models → Read.
- Direct API mode: No data stored; fetched from GitHub on each page load.
- Historical mode (optional): Cached in deployer-managed PostgreSQL for retention beyond 28 days.
- No telemetry, analytics, or phone-home. No data sent to publisher or third parties.
- AI conversations exist only in browser session — never stored server-side.
- Sessions use encrypted cookies (deployer-configured 32+ char secret).
- All API calls over HTTPS/TLS
- GitHub token server-side only, never exposed to browser
- Session encryption for auth state
- Content Security Policy headers
- No secrets in source code
- Self-hosted — publisher has no access to instances
- Open source for security audit
Report vulnerabilities via SECURITY.md.
Disabled by default. When enabled:
- Uses GitHub Models API for natural language Q&A about dashboard metrics
- AI sees ONLY metrics already visible on screen — no repo/code access
- Does not take actions or generate deployable code
- Conversations not persisted; discarded on page close
- Clearly labeled as AI-powered
| Service | Purpose | When |
|---|---|---|
| GitHub REST API | Copilot metrics & seats | Always |
| GitHub OAuth | User auth | OAuth mode only |
| GitHub Models API | AI chat | When enabled |
No other external services, analytics, or tracking.
Dashboard: Data visualization tool — not an AI system, outside EU AI Act scope.
AI chat (optional): Conversational data assistant. No biometric ID, social scoring, or automated decisions. Minimal risk (Articles 6, 8-17).
Published per the GitHub Marketplace Developer Agreement. Minimum permissions requested. Data practices described accurately above.
- Security: SECURITY.md
- Bugs: Issues
- Questions: Discussions