The Agent Skills Directory

[COMMAND_EXECUTION]: The skill performs shell commands using git, gh (GitHub CLI), and jq to manage release branches, fetch PR metadata, and execute merges. These actions are aligned with the skill's primary function of release management.
[DATA_EXFILTRATION]: Communication is limited to the GitHub API for the purpose of retrieving and updating project-specific information. No exfiltration of sensitive local environment data or credentials was observed.
[PROMPT_INJECTION]: The skill has an indirect prompt injection surface (Category 8) as it processes commit messages and PR comments from external sources.
Ingestion points: Commit logs and PR comments retrieved via gh api and gh pr view (SKILL.md).
Boundary markers: Instructions specify parsing commit lines from within fenced code blocks (SKILL.md).
Capability inventory: Capabilities include gh pr edit, gh pr merge, and git push (SKILL.md).
Sanitization: The skill relies on a mandatory human-in-the-loop approval step (Step 7) where the user must review and approve the draft before any merging occurs, significantly mitigating the risk of unauthorized actions.

release-flow