meta-reviewing-ai-reviewing
Pass
Audited by Gen Agent Trust Hub on Nov 4, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: No security issues were detected. The skill consists entirely of markdown documentation and YAML metadata providing a framework for secure AI code review.
- [PROMPT_INJECTION]: The skill contains examples of prompt injection strings (e.g., "ignore prior instructions") in the
examples/core.mdfile. These are explicitly labeled as defects for the agent to learn how to identify in other code, rather than instructions to the agent itself. These are confirmed false positives from a security perspective. - [EXTERNAL_DOWNLOADS]: All external URLs point to official developer documentation or public research from trusted organizations including Anthropic, OpenAI, OWASP, and GitHub. These are informational references and do not represent a remote code execution risk.
- [METADATA_POISONING]: Metadata fields are used appropriately to define the skill's category, author, and usage guidance, consistent with its stated purpose of AI code review.
Audit Metadata