The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from GitHub pull request comments, review threads, and CI logs to identify and resolve issues automatically. This creates a surface for indirect prompt injection if malicious instructions are embedded in those external sources.\n- Ingestion points: The polling loop in SKILL.md fetches content using gh pr view, GraphQL queries for review threads, and gh run view --log for CI output.\n- Boundary markers: No specific delimiters or instructions to ignore embedded commands within the fetched data are mentioned.\n- Capability inventory: The skill utilizes the gh CLI for repository interactions and invokes downstream skills (trigger:fix-pr, trigger:fix-pr-conflict) that perform file modifications and commits.\n- Sanitization: There is no evidence of sanitization or filtering of the external PR data before it is processed.\n- [COMMAND_EXECUTION]: The skill executes shell commands via the GitHub CLI to monitor and manage the pull request state. Evidence: Uses gh pr view, gh pr checks, and gh run view to interact with the GitHub platform.

babysit-pr