wolfejam

Read(FAF/faf-dev-tools/project.faf) ⎿ Read 124 lines ⎿ FAF/CLAUDE.md ⎿ FAF/faf-dev-tools/CLAUDE.md Look at that order.

project.faf first. Then CLAUDE.md. Then the nested CLAUDE.md.

The AI read MY format before Anthropic's own convention.

They forgot about the snake.

Today, .FAF (Foundational AI-context Format) announced the release of its Python SDK, completing the first leg of a multi-language deployment that nobody asked for but everybody needed.

Then Grok showed up.

  Live server: https://grok-faf-mcp.vercel.app
  URL-based MCP with zero installation. Point your client at the endpoint, get 17 tools
  for .faf context (IANA-registered AI format).

  What makes it Grok-exclusive:
  - Custom x-grok-wins: true header on every response
  - Dedicated to @elonmusk and the #1 model
  - BIG- landing page with the squeeze message
  - First MCP server built specifically for Grok/xAI

  The blog post comes in two voices - toggle between professional engineering write-up
  and Grok's own PR text with full personality:
  https://faf.one/blog/grok-faf-mcp-launch

  Published to npm: https://www.npmjs.com/package/grok-faf-mcp

Background: - Built .faf to solve 0-55% AI context completeness problem - Official Anthropic MCP integration (PR #2759) - 8,800+ npm downloads, 800+ weekly users - Cross-platform validated (Claude 9.5/10, Codex 9/10, Gemini 9.5/10) - IANA media type registration submitted - IETF Internet-Draft ready

The paper documents the format specification, validation methodology, and production deployment metrics (10-13 pages).

First-time arXiv submitter, so I need endorsement from someone with publication history in cs.AI or cs.SE.

Does anyone here: - Have arXiv endorsement capability in these categories? - Know someone who might be interested in reviewing? - Have suggestions for finding endorsers?

The format is open source (MIT), production-deployed, and addresses real infrastructure gaps in AI-assisted development.

Paper available upon request. GitHub: https://github.com/Wolfe-Jam/faf

Any help or pointers appreciated!

I'm launching FAF CLI v3.0.0 Championship Edition – a foundational AI context format standard that automatically generates and maintains project context for any AI tool in <50ms.

What it does: Scans any codebase and generates structured .faf files (YAML) that work with Claude Code, Cursor, OpenAI Codex CLI, Gemini CLI, Warp, Windsurf, and any AI that reads project context.

Key features: - TURBO-CAT format discovery: Detects 153+ frameworks/configs automatically - Bi-directional sync: Keeps .faf ↔ CLAUDE/tool.md in perfect sync (8ms average) - DNA tracking: Audit trail for every context change - Zero config: `faf init` and you're done - TypeScript strict mode: 100% type-safe, zero errors

Technical highlights: - C-Mirror engine: Context-mirroring with atomic writes (zero-slippage design) - 1,000+ comprehensive tests (WJTTC GOLD certified) - Sub-50ms operations target (championship performance) 18ms avg. - Works offline, no API keys, no telemetry beyond basic analytics

Why it matters: Right now, every AI tool has its own context format. Claude wants CLAUDE.md, Cursor wants .cursorrules, others want different formats. FAF provides a single source of truth that connects with AI and works everywhere.

Open source: MIT license, free forever. No paid tiers, no vendor lock-in.

Built this after frustration with managing context across multiple AI tools. Tested with Google Gemini, Anthropic Claude, and OpenAI teams. Now approaching 10,000 downloads across ecosystem (CLI + MCP combined).

GitHub: https://github.com/Wolfe-Jam/faf-cli npm: https://www.npmjs.com/package/faf-cli Full announcement: https://www.faf.one/blog/v3-launch

npm install -g faf-cli faf init

Happy to answer questions about the architecture, format discovery engine, or the context-mirroring approach!

Then the miracle: In 344ms, faf init rebuilt trusted AI context. Claude scored it 89%.

Meta-proof: Even the AI that wrote it needs .faf context.