What problem does it solve? AI-facing repository documents (README.md, AGENTS.md, CLAUDE.md, onboarding pages) drift out of sync with the actual project state as scripts, skills, aliases, and paths change, causing AI assistants to load stale or contradictory instructions. ## Core Features & Use Cases - Multi-document drift patching: Audits all AI-critical files using five discovery rules, then patches only verified drifted facts while preserving human prose and protected policy sections. - Cross-document consistency checks: Verifies that skill names, script names, environment URLs, and paths agree across every document before writing anything. - Security redaction protocol: Scans drafted content for credentials, production URLs, and shaped secrets (JWTs, PATs, AWS keys) before any write, with a user-visible redaction log. - Use Case: After renaming a skill or changing package.json scripts, run the sync to update README script tables, the AGENTS.md skills registry, and the onboarding HTML in one pass without rewriting stable rule sections. ## Quick Start Ask the AI to run sync-ai-context on this repository to detect and patch documentation drift across all AI-critical files.