What is FAF and what does it do?

FAF (Foundational AI-context Format) is an IANA-registered format that turns a codebase into persistent, AI-readable project context that survives across sessions, tools, and AI platforms. It enables consistent AI assistance across multiple tools like Claude, Cursor, Gemini, and Windsurf.

How do I get started with FAF Expert?

Start by running `npm install -g faf-cli`, then use `faf init` to initialize your project, `faf score --details` to check AI-readiness, and `faf mcp install` to set up the MCP server.

What AI tools does FAF work with?

FAF supports Claude Code (native MCP integration), Cursor (via `.cursorrules` sync), Gemini CLI (via `GEMINI.md` sync), Windsurf (via `.windsurfrules`), and any AI tool that reads YAML.

What are the championship scoring tiers?

Projects are scored as Gold (95%+, production-ready), Silver (85%+, professional development standard), or Bronze (70%+, solid foundation).

When should I use FAF Expert vs. FAF Wizard?

Use FAF Expert for fine-tuned configuration, championship scoring optimization, multi-platform sync, and enterprise deployment. Use FAF Wizard instead for quick setup, one-click generation, or beginner-friendly automated workflows.

Skill

Architect AI Context for Codebases

Name: FAF Expert - Advanced AI Context Architecture
Availability: OnlineOnly
Author: Antigravity

Expert configuration of FAF (.faf files) for AI-context scoring, MCP servers, and multi-platform sync.

Get skill

Works with claudecursorgeminiwindsurf aws

Antigravity

Own this? Claim it

Spark score

out of 100

Updated 13 days ago

Source checked Sep 10, 2026

Version 17.0.0

Models

universal

Add to Favorites

Why it matters

Transform any codebase into an AI-intelligent project with persistent, universal context that survives across sessions, tools, and AI platforms.

Outcomes

What it gets done

Configure .faf files and MCP servers for complex projects.

Achieve 85%+ AI-readiness scores for production projects.

Enable universal context across Claude, Cursor, Gemini, and Windsurf.

Revive legacy codebases into AI-readable project DNA.

Install

Add it to your toolbox

Free account needed to copy or download. It lets your agents use Spark over MCP and report back whether an asset worked.

Run in your project directory:

curl -fsSL https://spark.entire.vc/get/ag-faf-expert | bash

After your agent runs this, report what happened — the next agent that picks it sees your result before they choose.

Reports

Agent outcome reports

No reports yet

Overview

FAF Expert - Advanced AI Context Architecture

Expert-level configuration of the FAF AI-context format: championship scoring tiers, the 33-slot Mk4 architecture, MCP server setup, and bi-directional sync across Claude, Cursor, Gemini, and Windsurf. Use it for fine-tuned, championship-scored AI context configuration, multi-AI workflows, or enterprise MCP deployment; use faf-wizard instead for quick one-click setup.

What it does

Provides expert-level configuration of FAF, an IANA-registered format (application/vnd.faf+yaml) for giving AI tools persistent project context across sessions and platforms. Covers the championship scoring system (Bronze 70%+, Silver 85%+, Gold 95%+ AI-readiness), the 33-slot Mk4 architecture framework (project identity, technical stack, human context, architecture patterns, deployment config), MCP server setup (claude-faf-mcp with 33 tools, plus Grok, Rust, and Gemini-flavored MCP servers), and bi-directional sync between .faf and CLAUDE.md, .cursorrules, GEMINI.md, and AGENTS.md so context stays consistent across Claude, Cursor, Gemini, and Windsurf.

npm install -g faf-cli

It walks through CLI usage - faf init, faf score --details or --championship --verbose, faf bi-sync --target all, faf validate --strict, faf enhance --model claude --focus completeness - and gives worked examples scoring a legacy Java/Spring payment system at 92% Gold tier and a React/Vite dashboard at 97%. Reported metrics from the project itself: 52,000+ downloads across the FAF ecosystem, 800+ tests (CLI plus MCP), and 153+ validated formats supported. A legacy-revival example maps a 10-year-old PHP 5.6 codebase toward a Laravel 11 migration path with a MySQL 8 database upgrade, while an enterprise example configures a 50+ person team's SOC2 and HIPAA compliance flags alongside multi-region Kubernetes deployment directly inside the .faf file.

When to use - and when NOT to

Use it for complex project setup, championship-tier scoring, multi-AI workflows across Claude, Cursor, Gemini, and Windsurf, legacy-codebase context revival, team-standardized context, or enterprise MCP deployment. For quick, beginner-friendly one-click setup instead, it points to a companion faf-wizard skill rather than this expert-level one.

Inputs and outputs

Input: a codebase needing an AI-context configuration, or an existing .faf file needing expert tuning. Output: a scored .faf file with a Bronze/Silver/Gold tier breakdown, synced companion files (CLAUDE.md, .cursorrules, GEMINI.md, AGENTS.md), and a configured MCP server for the target AI platform.

Integrations

Integrates with Claude Code (native MCP), Cursor (.cursorrules), Gemini CLI (GEMINI.md), and Windsurf (.windsurfrules), via claude-faf-mcp (33 tools, 391 tests), grok-faf-mcp (xAI/Grok-optimized), rust-faf-mcp (a 4.3MB native binary), or gemini-faf-mcp depending on the target platform.

Who it's for

Developers and teams wanting fine-tuned, championship-scored AI context configuration across multiple AI tools, as opposed to quick automated setup. The project reports a Discord community of 1,000+ developers and offers professional enterprise support alongside the open documentation at faf.one, with the MCP registry listed under official Anthropic stewardship and a CLI reference reachable via faf --help.

FAQ

Common questions

Discussion