pipeline-test-runner

To validate generated pipeline skills against real targets, you execute them using a validated pipeline spec JSON to discover test fixtures, run subdomain skills in parallel, and generate per-subdomain pass/fail reports.

Pipeline testing with parallel execution runs multiple subdomain skills simultaneously against discovered test targets like codebase files and synthetic inputs, enforcing per-subdomain timeouts to catch runtime issues missed by structural checks.

You test pipeline fixtures and synthetic inputs by running the test runner against a validated pipeline spec, which automatically discovers these test targets for each subdomain and executes the skills to measure runtime performance.

Yes, pipeline artifact validation checks dual-layer artifacts by validating manifest.json and content.md files using dedicated artifact validation scripts, generating per-subdomain traces to ensure generated pipelines produce expected outputs.

A dual-layer report for pipeline testing is a per-subdomain pass/fail output consumable by pipeline-retro, needed when you want to analyze runtime execution results and trace artifact validation failures after executing generated skills.