The Agent Skills Directory

[SAFE]: The skill consists of markdown instructions outlining a manual review workflow for engineering documents. No scripts, binaries, or command-line patterns representing a security threat were found.
[INDIRECT_PROMPT_INJECTION]: The skill functions by analyzing external inputs such as RFC drafts and originating request notes. While this exposes a surface for indirect prompt injection, the risk is negligible as the instructions focus on analytical checks and reporting findings to a human author rather than performing automated actions, code execution, or network exfiltration. Ingestion points: RFC files, originating issues, and notes (SKILL.md). Boundary markers: Not specified. Capability inventory: Retrieval and analysis of local files and documentation. Sanitization: None mentioned.

review-rfc