The Agent Skills Directory

[PROMPT_INJECTION]: The skill processes untrusted data from Pull Request comments and CI logs which could contain malicious instructions.
Ingestion points: Fetches PR metadata, review comments, and inline comments using gh pr view and gh api in SKILL.md. It also retrieves CI logs via gh run view.
Boundary markers: The skill does not define specific delimiters or guardrails to prevent the agent from obeying instructions embedded in the comments or logs.
Capability inventory: The skill can perform file modifications and execute git push to update the remote repository.
Sanitization: No evidence of sanitization or filtering of the ingested external text is present.
[COMMAND_EXECUTION]: The skill relies on executing system commands to perform its core functions.
Evidence: SKILL.md defines the use of gh (GitHub CLI) for viewing and checking PR status and git for committing and pushing code changes.

iterate-pr