# Privacy Policy **claude-faf-mcp** — Privacy Policy ## Summary FAF processes everything locally on your machine. No data leaves your computer. ## What FAF Does - Reads and writes `project.faf` files in your project directories - Syncs context between `.faf`, `CLAUDE.md`, `AGENTS.md`, `.cursorrules`, and `GEMINI.md` - Scores your project's AI-readiness based on local file content - Detects frameworks and languages by scanning local files ## What FAF Does NOT Do - **No external network requests** — all processing is local, with one exception: `faf_git` fetches public repository metadata from the GitHub API (`api.github.com`) when you explicitly ask it to generate context from a GitHub URL. No data is sent; it only reads public repo information. - **No analytics or telemetry** — we don't track usage - **No data collection** — nothing is sent anywhere - **No user accounts** — no authentication, no sign-up - **No cookies or local storage** — beyond the files you ask it to create - **No third-party services** — zero external dependencies at runtime (except GitHub API for `faf_git`, user-initiated only) ## File Access FAF only accesses files you explicitly point it to (via `path` parameters or your current working directory). It creates and modifies: - `project.faf` — your project's AI context - `CLAUDE.md` — Anthropic Claude instructions (via bi-sync) - `AGENTS.md` — OpenAI/Codex instructions (via export) - `.cursorrules` — Cursor IDE instructions (via export) - `GEMINI.md` — Google Gemini instructions (via export) No files are read or written outside your project directory without your explicit request. ## Open Source This project is MIT-licensed and fully open source. You can verify every claim in this policy by reading the source code: https://github.com/Wolfe-Jam/claude-faf-mcp ## Contact Questions about privacy: team@faf.one --- *Last updated: 2026-04-15*