The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses shell commands to interact with the repository and verify changes.
Evidence: Uses git diff, git show, and grep to determine scope and verify orphans.
Evidence: The 'Verify' section requires running the project's own lint and test commands to validate fixes.
[INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from the repository being reviewed, which could theoretically contain instructions designed to manipulate the agent's behavior during the review process.
Ingestion points: The skill reads data from git diff, git show, and specific file paths provided by the user or identified via Git.
Boundary markers: The instructions do not define explicit delimiters (e.g., XML tags or specific markdown blocks) to isolate the code being reviewed from the agent's core instructions.
Capability inventory: The agent is authorized to execute file system reads (git, grep), file writes (to fix defects), and general command execution (running the project's lint/test suite).
Sanitization: There is no mention of sanitizing or escaping the content of the diffs before the agent analyzes them.

code-review