The Agent Skills Directory

[Indirect Prompt Injection] (SAFE): The skill processes untrusted code changes from external pull requests, creating a surface for potential indirect prompt injection. 1. Ingestion points: The skill uses gh pr view and gh pr diff in SKILL.md to read PR data into the agent context. 2. Boundary markers: Absent; there are no specific markers or instructions to ignore instructions found within the code diffs. 3. Capability inventory: The skill can execute local commands (git, gh, pnpm). 4. Sanitization: Absent; the skill does not explicitly sanitize the diff output.
[Command Execution] (SAFE): The skill utilizes command-line tools to perform its primary function. Evidence: It executes git diff, git log, gh pr, and various pnpm lifecycle scripts (pnpm build, pnpm test). These are standard operations for code analysis and review.

code-review