What problem does it solve? AI coding agents often start implementation without the project-specific conventions, architecture, and domain constraints they need, producing code that ignores existing patterns. This Skill enforces a mandatory Phase 0 gate that locates and loads functional AGENTS.md context files before any clarifying questions or code changes begin. ## Core Features & Use Cases - Domain-Based Discovery: Infers the relevant domain from arguments, file paths, or task descriptions, then globs for matching *AGENTS.md files with prioritized exact, parent, and related-domain matching. - Blocking Gate with Options: When no context exists, it blocks execution and offers explicit choices — create a new file from the template, search more broadly, provide paths manually, or bypass. - Template-Based Creation: Generates minimal AGENTS.md files from .agents/templates/TEMPLATE_AGENTS.md, deliberately left sparse for later population in Phase 8. - Use Case: A developer asks the agent to add an endpoint to the orders API; the Skill detects the orders domain, loads ORDER_PROCESSING_AGENTS.md, summarizes it, and only then proceeds to requirement clarification. ## Quick Start Ask the agent to load the AGENTS context for the auth domain before starting the login feature implementation.
