bro-review-spec-plan
Pass
Audited by Gen Agent Trust Hub on Oct 27, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data in the form of specifications and plans, which are passed to subagents for evaluation.
- Ingestion points: Data enters through file paths, message text, or attached files as identified in SKILL.md.
- Boundary markers: The prompts in reviewer-prompt.md and security-reviewer-prompt.md interpolate artifact content without strong delimiters or specific instructions to ignore embedded commands.
- Capability inventory: The subagents are capable of reading repository files, searching code, and making web requests as defined in the READONLY section of SKILL.md.
- Sanitization: No sanitization or escaping is performed on the ingested content prior to subagent processing.
Audit Metadata