MCP Catalogs
Homefaf screenshot

faf

by Wolfe-Jam·15·Score 42

FAF is an IANA-registered format that provides structured AI context through project.faf files and MCP servers.

developer-toolsai-llm
3
Forks
1
Open issues
1 mo ago
Last commit
2d ago
Indexed

Overview

The Foundational AI-Context Format (FAF) is an IANA-registered open format that transforms project context into structured data that AI systems can instantly understand. It provides a universal, shareable context format for any AI tool, human, or team. The project includes MCP servers that enable Claude and other AI systems to access project context through the FAF format, which exists as a YAML file with detailed information about dependencies, architecture, and project goals. The ecosystem has over 27,000 downloads across multiple implementations.

Try asking AI

After installing, here are 6 things you can ask your AI assistant:

you:Provide AI coding assistants with structured project context through MCP servers
you:Convert project files into a standardized AI-optimized format
you:Generate AI-specific instructions (Claude, Cursor, Gemini, ChatGPT) from a single source
you:What is FAF?
you:How does FAF compare to other AI context files?
you:What AI platforms support FAF?

When to choose this

Choose FAF when working across multiple AI coding assistants and need a standardized format to share project context consistently.

When NOT to choose this

Avoid FAF if you only use a single AI coding assistant and prefer to maintain separate instruction files for each tool.

Tools this server exposes

6 tools extracted from the README
  • faf_init

    Initialize a new FAF project

  • faf_migrate

    Migrate legacy .faf files to visible project.faf

  • faf_git

    Analyze GitHub repository and create project.faf

  • faf_score

    Calculate context completeness score for FAF files

  • faf_bi-sync

    Generate multiple AI instruction formats from FAF

  • faf_convert

    Convert any file type to .faf format

Note: Tool names inferred from CLI examples and descriptions in the README. The repository contains the format specification but the actual MCP tools are provided by separate packages like claude-faf-mcp.

Comparable tools

agents-md-mcpcursor-rules-mcpclaude-config-mcp

Installation

Installation

CLI Tool (recommended for getting started)

npm install -g faf-cli
cd your-project
faf init

Claude Desktop Integration

npm install -g claude-faf-mcp

Add to Claude Desktop configuration (claude_desktop_config.json):

{
  "mcpServers": {
    "faf": {
      "command": "claude-faf-mcp",
      "args": []
    }
  }
}

FAQ

What is FAF?
FAF (Foundational AI-Context Format) is an IANA-registered open format for AI context that transforms project information into structured data that AI systems can instantly understand.
How does FAF compare to other AI context files?
Unlike markdown-based context files (like .cursorrules or CLAUDE.md), FAF uses structured YAML, is IANA-registered as an Internet standard, supports scoring, and can generate all other AI-specific formats from a single source.
What AI platforms support FAF?
FAF has MCP servers for Claude, and implementations for Grok, Gemini, and other platforms. It's designed to be universal and works with any AI coding assistant.

On Hacker News

Recent discussion from the developer community.

Compare faf with

GitHub →

Last updated · Auto-generated from public README + GitHub signals.