The Agent Skills Directory

[SAFE]: The skill defines a behavioral framework for receiving technical feedback, prioritizing codebase verification and technical reasoning over performative agreement. This approach inherently encourages the agent to validate external input before execution.
[COMMAND_EXECUTION]: The skill references the use of grep for searching the codebase and the gh (GitHub) CLI for responding to pull request comments. Both tools are used in a standard, task-appropriate manner for software development workflows.
[PROMPT_INJECTION]: The skill includes instructions to ignore or avoid specific conversational patterns (e.g., 'Forbidden Responses'). These are stylistic persona constraints designed for technical efficiency and do not constitute an attempt to bypass system safety guidelines or security filters.

receiving-code-review