no-comments
Pass
Audited by Gen Agent Trust Hub on Oct 15, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests and processes untrusted user-provided source code and diffs to automate comment removal and identify refactor targets.
- Ingestion points: User-provided files or diffs are used to create the context for the
delegate_tasksubagent (SKILL.md, Step 1). - Boundary markers: Absent. The skill does not employ delimiters to separate the untrusted code content from the instructions being followed by the agent.
- Capability inventory: The agent is authorized to perform file modifications, including deleting code paths, dropping parameters, and implementing root-cause fixes (SKILL.md, Steps 3 and 4).
- Sanitization: Absent. The logic specifically explicitly instructs the subagent to interpret and act upon text markers like
IMPORTANTordo not removefound within the processed code, which could be used as an injection vector (SKILL.md, Step 2; references/comment-sicko.md).
Audit Metadata