Agent skill

faf-expert

Expert in .faf (Foundational AI-context Format) files for persistent project context. Use when working with .faf files, project DNA, CLAUDE.md bi-sync, faf-cli commands, MCP server configuration, or AI-readiness scoring (0-100%). Updated for v2.8.0 Tool Visibility System.

View SKILL.md on GitHub Repository
Stars 26
Forks 4

Install this agent skill to your Project

npx add-skill https://github.com/Wolfe-Jam/faf-cli/tree/v6/skills/faf-expert

SKILL.md

FAF Expert - Project DNA Specialist

What is .faf?

.faf is the universal format for persistent AI project context - "Project DNA ✨ for ANY AI".

Core Concepts:

  • Foundational AI-context Format - YAML-based project context file
  • Persistent Context - Survives across sessions, tools, and AI assistants
  • Universal - Works with Claude, ChatGPT, Gemini, Cursor, any AI
  • Measurable - Championship scoring system (0-100% AI-readiness)
  • Official Anthropic MCP Steward - PR #2759 MERGED

When to Use This Skill

Activate this skill when:

  • Creating or updating .faf files
  • Working with CLAUDE.md files
  • Running faf CLI commands (init, auto, score, formats, bi-sync)
  • Scoring project AI-readiness
  • Setting up MCP server (claude-faf-mcp)
  • Configuring tool visibility (v2.8.0+)
  • Discussing project context or documentation

Key Files

  • .faf - YAML format, machine-readable project context
  • CLAUDE.md - Markdown format, human-readable project guide
  • Bi-Sync - .faf ↔ CLAUDE.md kept in sync automatically
  • project.faf - v1.2.0 standard naming

Common Commands

faf-cli Commands

bash
faf init              # Create .faf from project
faf auto              # Auto-detect stack and create .faf
faf score             # Rate AI-readiness (0-100%)
faf formats           # List 153+ supported formats
faf bi-sync           # Sync .faf ↔ CLAUDE.md
faf status            # Check project AI-readiness
faf validate          # Validate .faf structure
faf doctor            # Diagnose and fix issues
faf migrate           # Migrate to latest format

MCP Tools (claude-faf-mcp v2.8.0+)

v2.8.0 introduces Tool Visibility System:

  • 21 Core Tools (default) - Essential workflow
  • 51 Total Tools (opt-in) - All features

Configuration:

json
{
  "mcpServers": {
    "faf": {
      "command": "npx",
      "args": ["-y", "claude-faf-mcp@2.8.0"],
      "env": {
        "FAF_MCP_SHOW_ADVANCED": "false"
      }
    }
  }
}

Core Tools (21):

  • Essential Workflow: faf, faf_auto, faf_init, faf_innit, faf_status
  • Quality: faf_score, faf_validate, faf_doctor, faf_audit
  • Intelligence: faf_formats, faf_stacks, faf_skills, faf_install_skill
  • Sync: faf_sync, faf_bi_sync, faf_update, faf_migrate
  • AI: faf_chat, faf_enhance
  • Help: faf_index, faf_faq, faf_about

Advanced Tools (30+): Set FAF_MCP_SHOW_ADVANCED: "true" to enable:

  • Display variants: faf_display, faf_show, faf_check
  • Trust system: faf_trust, faf_trust_confidence, faf_trust_garage
  • File operations: faf_read, faf_write, faf_list, faf_exists, etc.
  • DNA tracking: faf_dna, faf_log, faf_auth, faf_recover
  • Utilities: faf_choose, faf_clear, faf_share, faf_credit

AI-Readiness Scoring

Championship Tiers:

  • 🏆 95-100% - Championship (Gold)
  • 🥇 85-94% - Podium (Silver)
  • 🥈 70-84% - Points (Bronze)
  • 🥉 55-69% - Midfield
  • 🟢 40-54% - Backmarker
  • 🟡 25-39% - Struggling
  • 🔴 10-24% - Critical
  • 🤍 0-9% - DNF (Did Not Finish)

Target: 85%+ for production projects

.faf File Structure

yaml
# Basic .faf structure
format: typescript-nextjs
language: typescript
framework: nextjs
version: "14.0.0"

structure:
  - src/
  - components/
  - pages/

dependencies:
  react: "^18.0.0"
  next: "^14.0.0"

testing: jest
deployment: vercel

# v1.2.0+ additions
project_name: "My Project"
ai_readiness_score: 87

Best Practices

  1. Start with faf auto - Let it detect your stack
  2. Review and refine - Auto-detection is 85% accurate, tweak as needed
  3. Run faf score - Target 85%+ for championship context
  4. Use faf bi-sync - Keep .faf and CLAUDE.md in sync
  5. Update regularly - As project evolves, update context
  6. Use core tools first - Advanced tools are opt-in for specific needs

The FAF Family

6 Championship Integrations:

  • React (20M/week npm downloads) 🏆
  • TypeScript (40M/week) 🏆
  • Next.js (5M/week) 🏆
  • Vite (9M/week) 🥇
  • Svelte (400K/week) 🥇
  • n8n (401K/week - TURBO only) 🥇

Installation

faf-cli (npm)

bash
npm install -g faf-cli

faf-cli (Homebrew)

bash
brew install faf-cli

claude-faf-mcp (MCP Server)

Add to ~/Library/Application Support/Claude/claude_desktop_config.json:

json
{
  "mcpServers": {
    "faf": {
      "command": "npx",
      "args": ["-y", "claude-faf-mcp@latest"]
    }
  }
}

For advanced tools:

json
{
  "mcpServers": {
    "faf": {
      "command": "npx",
      "args": ["-y", "claude-faf-mcp@latest"],
      "env": {
        "FAF_MCP_SHOW_ADVANCED": "true"
      }
    }
  }
}

v2.8.0 Tool Visibility System

Problem Solved: Reduced cognitive load from 51 tools to 21 core tools.

How It Works:

  • Default: Shows 21 essential tools
  • Opt-in: Enable all 51 tools via environment variable
  • Smart Filtering: Tools categorized by purpose
  • Backward Compatible: Existing setups continue working

Categories:

  • workflow - Essential commands
  • quality - Scoring and validation
  • intelligence - Format/stack detection
  • sync - Context synchronization
  • ai - AI enhancement features
  • help - Documentation
  • trust - Trust validation
  • file - File operations
  • utility - Misc tools

Configuration Priority:

  1. Environment variable (FAF_MCP_SHOW_ADVANCED)
  2. Config file (~/.fafrc)
  3. Default (core only)

Philosophy

The Noodle Philosophy 🍜

  • YAML noodles for AI (machine-readable)
  • Converts to markdown/TXT for humans (human-readable)
  • Interconnected context, not flat data

Format-First

  • The format is the foundation
  • Without .faf, there is no universal context
  • Format-driven architecture

Official Stewardship

  • Anthropic-approved MCP server (PR #2759 MERGED)
  • Account Managers for all things .FAF in Anthropic ecosystem
  • Governance responsibility for format specification

Brand Values

  • NO BS ZONE - Only real stats, verified claims
  • Championship Standards - <50ms performance, 1,000+ tests
  • Free Forever - MIT license, open source
  • Trust is Everything - Built on credibility
  • Professional, Boring, Trusted - F1-grade engineering

Stats (Real, Verified - November 2025)

  • 13,000+ total downloads (CLI + MCP combined)
  • 6,700+ MCP downloads (weekly: 800+)
  • 6,500+ CLI downloads
  • 79 tests passing (57 MCP + 22 visibility)
  • 153+ formats validated (TURBO-CAT engine)
  • <10ms tool filtering (5x better than 50ms target)
  • Zero regressions in v2.8.0 release
  • WJTTC Gold Certified - F1-inspired testing standards
  • Anthropic-approved MCP server (Official steward)

Version History

  • v3.2.7 - faf-cli latest
  • v3.3.7 - claude-faf-mcp latest (TYPE_DEFINITIONS parity)
  • v1.2.0 - faf-mcp universal

Resources

Testing Standards

WJTTC (WolfeJam Technical & Testing Center)

  • F1-Inspired testing methodology
  • 3 Tiers: Brake Systems, Engine Systems, Aerodynamics
  • Philosophy: "We break things so others never have to know they were broken"
  • All tests reported and documented
  • Gold certification for production releases

Troubleshooting

MCP Server Not Loading

bash
# Check logs
tail -f ~/Library/Logs/Claude/mcp-server-claude-faf-mcp.log

# Verify config
cat ~/Library/Application\ Support/Claude/claude_desktop_config.json

# Test locally
node /path/to/claude-faf-mcp/dist/src/index.js

Tool Count Verification

bash
# Should show 21 core tools (default)
# Or 51 tools (advanced mode)
# Check console output when MCP loads

CLI Not Found

bash
# Install globally
npm install -g faf-cli

# Or via Homebrew
brew install faf-cli

# Verify installation
faf --version

Advanced Usage

Custom Config File

Create ~/.fafrc:

json
{
  "showAdvanced": true
}

Or key=value format:

FAF_SHOW_ADVANCED=true

Local Development

bash
# Use local build
{
  "mcpServers": {
    "faf": {
      "command": "node",
      "args": ["/path/to/claude-faf-mcp/dist/src/index.js"],
      "env": {
        "FAF_MCP_SHOW_ADVANCED": "false"
      }
    }
  }
}

Skill Control

To disable this skill temporarily:

bash
# Rename directory to disable
mv ~/.claude/skills/faf-expert ~/.claude/skills/faf-expert.disabled

To re-enable:

bash
# Rename back to enable
mv ~/.claude/skills/faf-expert.disabled ~/.claude/skills/faf-expert

Or just delete the directory to remove completely.

How to Get This Skill

Option 1: Manual Installation

  1. Copy this SKILL.md file
  2. Create directory: ~/.claude/skills/faf-expert/
  3. Save as: ~/.claude/skills/faf-expert/SKILL.md
  4. Restart Claude Code

Option 2: Git Clone (if published)

bash
# Coming soon - published skills repository
git clone [skills-repo] ~/.claude/skills/faf-expert

Option 3: From FAF CLI (future)

bash
# Planned feature
faf skill install faf-expert

Made with 🧡 by wolfejam.dev Official Anthropic MCP Steward Championship Edition v2.8.0 🏎️✨ IANA-Registered Format: application/vnd.faf+yaml (Oct 31, 2025)

Recommended Agent Skills

Expand your agent's capabilities with these related and highly-rated skills.

davila7/claude-code-templates

verl-rl-training

Provides guidance for training LLMs with reinforcement learning using verl (Volcano Engine RL). Use when implementing RLHF, GRPO, PPO, or other RL algorithms for LLM post-training at scale with flexible infrastructure backends.

23,776 2,298
Explore
davila7/claude-code-templates

openrlhf-training

High-performance RLHF framework with Ray+vLLM acceleration. Use for PPO, GRPO, RLOO, DPO training of large models (7B-70B+). Built on Ray, vLLM, ZeRO-3. 2× faster than DeepSpeedChat with distributed architecture and GPU resource sharing.

23,776 2,298
Explore
davila7/claude-code-templates

gguf-quantization

GGUF format and llama.cpp quantization for efficient CPU/GPU inference. Use when deploying models on consumer hardware, Apple Silicon, or when needing flexible quantization from 2-8 bit without GPU requirements.

23,776 2,298
Explore
davila7/claude-code-templates

Claude Code Guide

Master guide for using Claude Code effectively. Includes configuration templates, prompting strategies "Thinking" keywords, debugging techniques, and best practices for interacting with the agent.

23,776 2,298
Explore
davila7/claude-code-templates

qdrant-vector-search

High-performance vector similarity search engine for RAG and semantic search. Use when building production RAG systems requiring fast nearest neighbor search, hybrid search with filtering, or scalable vector storage with Rust-powered performance.

23,776 2,298
Explore
davila7/claude-code-templates

behavioral-modes

AI operational modes (brainstorm, implement, debug, review, teach, ship, orchestrate). Use to adapt behavior based on task type.

23,776 2,298
Explore

Didn't find tool you were looking for?

Be as detailed as possible for better results