What problem does it solve? Setting up the blueprint-harness orchestration engine in a repository involves copying engine files, seeding configuration, wiring skills into AI assistant setups, and injecting wrapper blocks — a repetitive, error-prone process that this Skill automates with strict write boundaries and update detection. ## Core Features & Use Cases - Engine installation and updates: Copies the orchestrator, skills, schemas, and templates into .bp-harness/, comparing VERSION files to detect fresh installs, no-op repairs, or updates. - AI setup integration: Detects Claude Code (CLAUDE.md/.claude/) and AGENTS (AGENTS.md/.agents/) environments, copies bp-* skills into them, and injects marked wrapper blocks with user confirmation. - Workspace bootstrapping: Creates bp-workspace/ with objectives, ideas, bugs, and audits directories, seeds a schema-validated config.yaml, and detects capabilities like rg and gh. - Use Case: A developer clones a project and wants spec-driven development tooling; they run the installer once and get the engine, workspace, and AI assistant wrappers configured with at most two questions. ## Quick Start Ask the AI to run bp init to install blueprint-harness into the current repository.