The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses the GitHub CLI (gh) to perform repository operations. It includes comprehensive pre-flight checks and validation steps (e.g., verifying labels and milestones exist) before any mutation occurs.
[DATA_EXFILTRATION]: A dedicated 'Public-Repo Leak Guard' is implemented across multiple workflows (new-issue, write-changelog). This guard scans draft content for private repository URLs and file paths, refusing to proceed if a leak to a public repository is detected.
[PROMPT_INJECTION]: The skill processes external data (issue and pull request bodies) which represents an indirect prompt injection surface. However, the skill mitigates this risk through structured processing, explicit validation of inputs, and a refusal to perform cross-repository operations by default.

using-github