The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes git commands to retrieve repository history, resolve diffs, and perform commits. Evidence: The scripts/resolve_diff.py script uses subprocess.run to call git log and git diff with user-provided arguments. The SKILL.md orchestrates the execution of git add -A and git commit as part of its automated workflow.\n- [PROMPT_INJECTION]: The skill presents an indirect prompt injection vulnerability surface by processing external data (git diffs) that may contain malicious instructions designed to influence the agent's analysis. \n
Ingestion points: Raw git diff strings are fetched by resolve_diff.py and parsed into hunks for the agent to review.\n
Boundary markers: The references/review-format.md instructions dictate that the diff content should be displayed within markdown code blocks using diff syntax highlighting to separate it from instructions.\n
Capability inventory: The skill has the ability to modify project source code using the Edit tool (via the --apply command) and to execute commits to the repository (via the --commit command).\n
Sanitization: The skill does not implement any sanitization or filtering of the diff content to prevent the agent from interpreting embedded instructions as legitimate commands.

interactive-diff-review