The Agent Skills Directory

[PROMPT_INJECTION]: The skill ingests untrusted data from GitHub PR diffs, titles, and existing bodies via gh pr view and gh pr diff in SKILL.md. There are no explicit boundary markers or content sanitization for this data in the instructions. However, this interaction is a core part of its PR documentation function and the skill's write capability is scoped to the gh pr edit command.\n- [COMMAND_EXECUTION]: The skill uses the gh command-line tool. It instructs the agent in SKILL.md to use gh pr edit --body-file - and pipe the content, which is a secure practice to prevent shell character injection issues.

rjv-pr-descriptions