v1.2.2

What's new

Added

• CONTRIBUTING.md — TL;DR for xAI devs, PR conventions, branch model, code style, CI doctrine, MCP-tool contribution path, npm publish discipline, architecture decisions (.faf here / .fafm in the Voice variant). Tight (~170 lines), F1-inspired tone. Direct line for high-signal contributors landing on the repo.
• Voice variant cross-link in README — points readers at grok-faf-voice, the reference implementation of the Voice Memory Layer (VML). Same family (.faf 🐘 for context / .fafm 🐘🎙️ for voice memory), different surface.

Changed

• README H1 simplified — # grok-faf-mcp | FAST⚡️AF → # FAST⚡️AF Context. npm/PyPI both render the package name above the README, so the prefix doubled up. New H1 pairs with grok-faf-voice's "Fast⚡️AF memory setup" as sibling FAST⚡️AF taglines (.faf = Context, .fafm = memory).
• Tier ladder canonical alignment — README scoring table now matches faf-cli/src/core/tiers.ts exactly: 🏆 ★ ◆ ◇ ● ● ○ ♡. Adds the WHITE 0% row that was previously missing.
• README test count corrected: 179 → 169 (actual passing count). Honest pass for v1.2.2.

Notes

• Docs-only patch — no behavior changes, no MCP protocol changes, no new dependencies. Safe upgrade for all consumers.
• Companion to grok-faf-voice v0.1.3 launch (PyPI live as of 2026-04-30); the cross-link routes existing grok-faf-mcp traffic to the new Voice surface.

---

169 tests passing · MCP Registry: schema 2025-12-11 · Live on Vercel Edge

FAF defines. MD instructs. AI codes.