What problem does it solve? Setting up and maintaining a swamp repository involves many moving parts: initializing the directory structure, enrolling AI agent tools, configuring datastores, resolving stuck locks, and wiring up CI. This Skill consolidates all repository lifecycle operations into guided CLI workflows with predictable JSON output. ## Core Features & Use Cases - Repository Initialization & Upgrade: Create new repos with swamp repo init, enroll multiple AI tools (Claude Code, Kiro, Cursor, and more) with repeatable --tool flags, and upgrade existing repos with replace-semantics tool management. - Datastore Management: Check status, migrate between filesystem and extension datastores (e.g., S3), sync remote data, and force-release stuck distributed locks. - Extension Sources: Load extensions from external filesystem paths via .swamp-sources.yaml with glob support and type filters, overriding pulled registry extensions during local development. - Use Case: A team wants shared runtime data in S3. Run swamp datastore setup extension @swamp/s3-datastore with bucket config, verify health with swamp datastore status, and set SWAMP_DATASTORE in CI so every runner syncs automatically. ## Quick Start Ask the agent to initialize a new swamp repository in the current directory and enroll both Claude Code and Kiro as tools.