The Agent Skills Directory

[PROMPT_INJECTION]: The skill provides instructions that modify the agent's conversational style, specifically forbidding 'performative agreement' (e.g., 'You're absolutely right!') and gratitude. These are behavioral constraints designed to ensure professional technical rigor rather than attempts to bypass safety filters or ignore system instructions.
[PROMPT_INJECTION]: Indirect Injection Analysis:
Ingestion points: The skill is designed to process 'Code Review Feedback' from external reviewers (SKILL.md).
Boundary markers: The instructions contain explicit safeguards, requiring the agent to 'Verify before implementing', check if suggestions break functionality, and push back against incorrect advice.
Capability inventory: The skill utilizes grep for codebase searching and gh api for interacting with GitHub pull request comments.
Sanitization: The primary sanitization mechanism is the mandatory 'Technical Verification' step where the agent must check suggestions against 'codebase reality' before taking action.
[COMMAND_EXECUTION]: The skill references the use of grep to search the codebase for feature usage (YAGNI check) and gh api to post replies to specific GitHub comment threads. These commands are standard for a development-focused agent and are used here to facilitate technical evaluation and communication within a pull request workflow.
[EXTERNAL_DOWNLOADS]: The license file (LICENSE.txt) contains a reference to an external GitHub repository (github.com/obra/superpowers) for attribution purposes. This is a static reference and does not involve downloading or executing remote content during the skill's operation.

receiving-code-review