What problem does it solve? Orchestrators in an SDD v2 workflow need a fast, durable way to discover which canonical skills exist and what project rules apply, without re-reading every skill file and config document on each run. ## Core Features & Use Cases - Registry Generation: Scans canonical SDD v2 skills, classifies them as core, support, or shared-contract, and writes .sdd/skill-registry.md with paths, responsibilities, and stage readiness. - Compact Rules Derivation: Extracts session-scoped facts such as artifact root, package manager, quality commands, and language preference from openspec/config.yaml and .sdd/project-map.md. - Refresh Signal Tracking: Records lockfiles, tool configs, and key source directories so later runs know when bootstrap should be refreshed. - Use Case: During sdd-init, when bootstrap is missing or skill resolution previously failed, run this Skill to rebuild the registry so the orchestrator can route phases to the correct SDD skills. ## Quick Start Run the skill-registry skill to rebuild .sdd/skill-registry.md from the canonical SDD v2 skills and current project configuration.