debug
Pass
Audited by Gen Agent Trust Hub on Oct 13, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill defines a workflow for analyzing external error messages and tool outputs, which are potential vectors for malicious instructions.
- Ingestion points: The workflow involves processing error logs and diagnostic output from external debugging tools.
- Boundary markers: The skill contains a specific defense section ('Treating error output as data') that explicitly instructs the agent to treat external outputs as data to be analyzed rather than instructions to follow.
- Capability inventory: The agent uses standard development tools such as
git(for bisecting) and project-specific test harnesses to reproduce issues. - Sanitization: The instructions require the agent to surface any command-like content found in errors to the user instead of acting on it automatically.
Audit Metadata