The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from GitHub PR comments and reviews, using the content as direct instructions for parallel sub-agents.
Ingestion points: The skill fetches external content using gh pr view and gh api repos/{owner}/{repo}/pulls/PR_NUMBER/comments in SKILL.md.
Boundary markers: No explicit delimiters or instructions are used to separate the untrusted comment text from the agent's internal task instructions.
Capability inventory: The skill possesses the ability to spawn sub-agents (Task general-purpose(comment)), write to the filesystem (docs/conventions/), and perform git operations (git add, git commit, git push).
Sanitization: There is no evidence of sanitization or filtering of the comment content before it is passed to the sub-agent task environment.
[COMMAND_EXECUTION]: The skill utilizes shell commands through the GitHub CLI (gh) and Git to manage PR status and update documentation.
Evidence: Commands such as gh pr status, gh pr view, and gh api are used to retrieve metadata, and git commit is used to persist changes to the repository.

resolve-pr-parallel