ketchup
Pass
Audited by Gen Agent Trust Hub on Nov 3, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONDYNAMIC_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes external data from chat history, subagent reports, and CI status tools which could contain malicious instructions.
- Ingestion points: Chat messages from other users and subagent reports are used as primary input for the summary (SKILL.md).
- Boundary markers: The skill provides strong instructional boundaries, specifically stating 'Keep this turn read-only' and 'do not change code, start an app or a job, or post anything.'
- Capability inventory: The agent may use the
evaltool for data visualization and generate visual diagrams. - Sanitization: The instructions include requirements to cross-check numbers against source facts and verify resource states before reporting.
- [DYNAMIC_EXECUTION]: The skill instructs the agent to use the
evaltool to render charts from numerical data if a plotting library is available.
Audit Metadata