bug-investigator
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: No malicious patterns or behaviors were detected in the skill instructions or methodology. The skill provides a standard structured approach to technical debugging.
- [INDIRECT_PROMPT_INJECTION]: The skill instructions define an ingestion surface for untrusted data as part of its debugging methodology, which is an inherent part of technical root-cause analysis.
- Ingestion points: The skill directs the agent to read error messages, stack traces, and git diff outputs (SKILL.md, Phase 1).
- Boundary markers: None identified; the skill does not specify the use of delimiters or instructions to ignore potential commands within the ingested logs.
- Capability inventory: The skill utilizes the ssf CLI tool suite for state management, recording debugging attempts, and architectural escalation (SKILL.md, DP-5).
- Sanitization: No explicit sanitization or validation of the input error messages or stack traces is described before the agent processes them.
Audit Metadata