The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses the GitHub CLI (gh) and Git to fetch data and push changes, and it executes ruff and pytest locally. These commands are executed based on context and parameters derived from external pull request review data.
[REMOTE_CODE_EXECUTION]: The skill follows a workflow where code suggestions extracted from pull request comments are directly applied to the filesystem and subsequently executed during the pytest verification phase. This effectively allows an external source to dictate code that is executed on the runner.
[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection through its ingestion of external instructions. 1. Ingestion points: Pull request review comments fetched via the GitHub API (Step 2 in SKILL.md). 2. Boundary markers: No delimiters or explicit instructions to ignore embedded commands are present in the processing logic for comment bodies. 3. Capability inventory: The skill can modify files, execute code via pytest, perform linting/formatting with ruff, and push code to the repository via git push. 4. Sanitization: No sanitization, validation, or sandboxing of the suggested code or natural language instructions is performed before they are applied or executed.

fix-review