The Agent Skills Directory

[COMMAND_EXECUTION]: The skill invokes local git commands and project-specific development tools to analyze and modify the codebase.
Executes git diff, git status, and git commit to identify and manage code changes (SKILL.md, Steps 1 and 8).
Runs project-defined formatting and linting commands such as format:check and lint (SKILL.md, Step 2).
Applies fixes by writing generated code directly to the local filesystem and staging the changes (SKILL.md, Step 7).
[PROMPT_INJECTION]: The skill exposes a surface area for indirect prompt injection as it processes untrusted content from the codebase.
Ingestion points: The workflow reads arbitrary code content and diffs from the repository using git diff (SKILL.md, Step 1).
Boundary markers: The instructions lack explicit delimiters or "ignore embedded instructions" warnings for the code content being analyzed.
Capability inventory: The agent has the capability to write to the filesystem (applying fixes) and commit changes to the repository (SKILL.md, Steps 7 and 8).
Sanitization: There is no evidence of sanitization or filtering for instructions that might be embedded in code comments or string literals within the reviewed files.

review-code