problem-analysis
Pass
Audited by Gen Agent Trust Hub on Oct 8, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to investigate code, configurations, and documentation to identify root causes. This ingestion of external, potentially untrusted data constitutes an indirect prompt injection attack surface. If the files being analyzed contain malicious instructions, they could potentially influence the agent's behavior.
- Ingestion points: Local source code, configuration files, and documentation examined during the 'Investigate' phase.
- Boundary markers: The skill instructions do not define specific delimiters or warnings to separate processed data from agent instructions.
- Capability inventory: The agent follows instructions from an internal Python script (skills.problem_analysis.analyze) which may perform file reads and command execution.
- Sanitization: No sanitization, escaping, or validation of external content is described in the markdown documentation.
Audit Metadata