What problem does it solve?
This Skill eliminates the hassle of manually tracking, comparing, and cleaning up past AI analysis pipeline runs, which are often scattered across unorganized directories with no standardized metadata or easy way to access their results and status.
Core Features & Use Cases
- Run Listing & Inspection: View all past pipeline runs sorted by date, with key metadata including dataset, status, agent completion count, and timing. Drill down into individual runs to see agent status, output files, and confidence grades.
- Run Comparison: Side-by-side comparison of two runs to identify differences in metrics, findings, chart generation, and duration, ideal for evaluating changes to analysis configurations or datasets.
- Safe Run Cleanup: Delete runs older than 30 days with explicit user confirmation to free up storage, with automatic cleanup of associated symlinks if the latest run is deleted.
- Use Case: A data analyst running multiple churn analysis pipelines for different customer segments can use this Skill to quickly compare the latest two runs to see which model configuration performed better, or delete failed runs from 2 months ago to save disk space.
Quick Start
Use the manage-runs skill to display a sorted list of all your past pipeline runs with their status and completion details.
