Gemini FAF
Gives Gemini a persistent project.faf context file so sessions start with your stack
- Pricing
- Open source
- Platforms
- CLI, self-hosted
- Runs
- Locally
- Version
- 3.3.1
- Listing updated
- 3 days ago
About Gemini FAF
Gemini FAF is an open-source MCP server for ai models and llm tooling. It runs locally on your machine and is started by your AI client.
Every new Gemini conversation starts without knowing your stack or conventions. This server reads a single structured YAML file, project.faf, so Gemini begins each session already briefed. The README presents it as a Python FastMCP package on PyPI, with the .faf format registered with IANA.
The workflow splits three jobs. The .faf file defines the project, instruction files such as CLAUDE.md, AGENTS.md and .cursorrules can be written from it, and any AI then reads the same source. Failures such as a missing file, invalid YAML, a refused write or a path outside the project come back to the client as MCP errors.
Version 3.3.0 adds installation as a plugin for Antigravity CLI (agy), with /setup, /score and /export skills, and the server finds and writes project.faf in the folder you ran it from. The README names a faf_score tool that reports when a file is not valid YAML.
Install
Run in a terminal
claude mcp add gemini-faf-mcp -- uvx gemini-faf-mcpAdd to claude_desktop_config.json
{
"mcpServers": {
"gemini-faf-mcp": {
"command": "uvx",
"args": [
"gemini-faf-mcp"
]
}
}
}Add to ~/.cursor/mcp.json
{
"mcpServers": {
"gemini-faf-mcp": {
"command": "uvx",
"args": [
"gemini-faf-mcp"
]
}
}
}Add to .vscode/mcp.json
{
"servers": {
"gemini-faf-mcp": {
"type": "stdio",
"command": "uvx",
"args": [
"gemini-faf-mcp"
]
}
}
}Add to ~/.codeium/windsurf/mcp_config.json
{
"mcpServers": {
"gemini-faf-mcp": {
"command": "uvx",
"args": [
"gemini-faf-mcp"
]
}
}
}Add to your client's MCP settings (most use this shape)
{
"mcpServers": {
"gemini-faf-mcp": {
"command": "uvx",
"args": [
"gemini-faf-mcp"
]
}
}
}What it does
- project.faf context
- One YAML file holds the stack, goals and conventions the model should know.
- Python FastMCP server
- Runs locally over stdio from the gemini-faf-mcp package on PyPI.
- Antigravity plugin
- Installs in the agy CLI and works in the folder where agy was started.
- Setup, score and export skills
- Skills /setup, /score and /export manage the file (/faf:setup in Gemini CLI).
- Real error reporting
- Missing files, invalid YAML and unsafe paths return MCP isError results.
Good for
- Brief Gemini on a new codebase once instead of every session
- Generate AGENTS.md or CLAUDE.md from one source file
- Score how complete a project.faf file is
- Keep several AI tools reading the same project context
Pricing
Open source
Prices change. Check faf.one before you buy.
Gemini FAF FAQ
What does the server read?
A project.faf file in your project folder.
How is it installed?
From PyPI as gemini-faf-mcp, or in Antigravity CLI with agy plugin install and the repository URL.
Does it write files?
Yes, it can write project.faf in your project, and the README says it never writes to a folder broader than the project.
What transport does it use?
stdio.
Is Gemini FAF free?
Yes. Gemini FAF is open source, so you can use and self-host it for free.
How do I add Gemini FAF to Claude Code?
Run this command in your terminal: claude mcp add gemini-faf-mcp -- uvx gemini-faf-mcp
What are the best alternatives to Gemini FAF?
The closest alternatives to Gemini FAF are REA, Codebase Memory, GitHub, Serena MCP: the IDE for your agent and XERJ.
Alternatives to Gemini FAF

REA
MCP serverCLI and MCP server that lets coding agents reverse engineer apps and binaries
Runs locally- Open source
- 464K GitHub stars
- Developer tools

Codebase Memory
MCP serverLocal MCP server that builds a code knowledge graph for AI coding agents
Runs locally- Open source
- 48.7K GitHub stars
- Knowledge and memory

GitHub
MCP serverOfficial MCP server connecting AI assistants to GitHub repos, issues and workflows
Not checked yet- Remote
- Open source
- 34.5K GitHub stars
- Developer tools

Serena MCP: the IDE for your agent
MCP serverMCP toolkit giving coding agents symbol-level code retrieval and editing
Runs locally- Open source
- 31.5K GitHub stars
- Developer tools

XERJ
MCP serverLocal search engine for AI agents with cited answers over your code and docs
Runs locally- Open source
- 19.2K GitHub stars
- Knowledge and memory

Ouroboros
MCP serverSpec-first layer that pins requirements and verifies results of AI coding agent work
Runs locally- Open source
- 6.4K GitHub stars
- Developer tools