The Agent Skills Directory

[SAFE]: The skill consists entirely of markdown-based instructional content. No executable scripts, command-line operations, or network requests are present.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process untrusted external code and diffs (Ingestion points: code/diff provided by user). While it does not specify explicit input delimiters (Boundary markers), it follows a rigorous five-step analytical workflow and a fixed output schema. The skill does not utilize any tools or perform network, file, or system operations (Capability inventory: none), and no specific sanitization logic is implemented for the processed code (Sanitization: absent). Due to the lack of dangerous capabilities, the risk is negligible.

code-review