pipeline-health-report
Pass
Audited by Gen Agent Trust Hub on Jun 20, 2026
Risk Level: SAFE
Full Analysis
- [COMMAND_EXECUTION]: The skill leverages a specific set of tools (search_workflows, get_latest_instance, get_instance_failures, get_instance_summary) through the Kragle MCP to gather pipeline health metrics. This is controlled, purposeful tool usage consistent with the skill's stated intent.
- [DATA_EXFILTRATION]: The skill references a shared pattern for uploading HTML reports to S3. This is a standard delivery mechanism for report generation within the defined platform context and does not exhibit malicious exfiltration patterns.
- [INDIRECT_PROMPT_INJECTION]: The skill processes external data (API responses containing workflow statuses and error messages) which are then interpolated into a report. Although this is an ingestion surface for untrusted data, the skill includes strict instructions for 'deterministic math' and 'no invented data,' requiring the agent to narrate existing numbers rather than interpret instructions from the data, which mitigates injection risks.
Audit Metadata