sub-agent-review
Pass
Audited by Gen Agent Trust Hub on Oct 1, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and evaluate sub-agent configuration files located at user-provided paths ($ARGUMENTS) or standard directories (~/.claude/agents/). These files are untrusted external data.
- Ingestion points: Files at paths specified in $ARGUMENTS and standard agent directories (~/.claude/agents/).
- Boundary markers: The instructions do not define specific delimiters or instructions to ignore embedded prompts within the files being reviewed.
- Capability inventory: The skill relies on the agent's existing capabilities (e.g., file reading tools) to perform the review.
- Sanitization: No specific sanitization logic is provided for the ingested sub-agent content.
Audit Metadata