debugging-and-error-recovery
Pass
Audited by Gen Agent Trust Hub on May 7, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: No malicious patterns or security risks were identified. The content is purely instructional and focused on software debugging best practices.- [PROMPT_INJECTION]: The skill proactively addresses indirect prompt injection risks by instructing the agent to treat external data, such as stack traces and error output, as untrusted and to avoid following embedded instructions or URLs found within diagnostic data.
Audit Metadata