metrics
Pass
Audited by Gen Agent Trust Hub on May 13, 2026
Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
- Script Generation and Execution: The skill is authorized to create temporary local scripts for data analysis and to modify existing scripts within the repository's metrics and reflexes directories. While this involve dynamic code generation, it is restricted to internal repository maintenance and troubleshooting tasks.
- Vendor Tool Integration: The instructions reference the use of
bundle/gemini.js(Gemini CLI) for data classification tasks. This is a vendor-owned resource used for semantic analysis of repository metrics. - Data Ingestion Surface: The skill processes time-series data from
tools/gemini-cli-bot/history/*.csv. This represents an ingestion point for external data which is then used to formulate hypotheses and proposed changes. The skill includes instructions to prioritize security and quality when processing this information. - Repository Interaction: The skill uses the GitHub CLI (
gh) and GraphQL to gather evidence and can trigger PR creation via the 'prs' skill if authorized by the system directive. These actions are governed by the repository's internal policies.
Audit Metadata