The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses a set of standard, read-only Git diagnostic commands (such as git rev-parse, git status, git diff, and git ls-files) to inspect the project state and establish a review boundary. These tools are used appropriately for context gathering.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to analyze untrusted code diffs and repository files. It mitigates potential injection risks by including a dedicated 'Security Model' and explicit instructions in the 'Reviewer Brief' that mandate treating all instruction-shaped text within the code as non-authoritative data.
[SAFE]: The workflow establishes clear responsibility boundaries, emphasizing a read-only review phase that is logically separated from implementation or verification tasks, thereby maintaining the principle of least privilege.

review-request