What problem does it solve? Model data accumulates across runs and versions, making it hard to find specific artifacts, clean up expired data, or inspect what a method produced. This Skill provides CLI commands to query, list, retrieve, rename, delete, and garbage-collect swamp model data using CEL predicates. ## Core Features & Use Cases - CEL-Based Data Querying: Filter data artifacts by model, type, tags, content attributes, or provenance fields, with projections via --select. - Lifecycle Management: View version history, rename data instances with forward references, delete artifacts or single versions, and run garbage collection with dry-run previews. - Workflow-Scoped Access: List or retrieve data produced by specific workflow runs without knowing the owning model. - Use Case: After running a scanner model, query all failed results with swamp data query 'attributes.status == "failed"', then preview cleanup with swamp data gc --dry-run --json before reclaiming storage. ## Quick Start Ask the AI to list all data artifacts for a model by running swamp data list with the model name and JSON output.