The Agent Skills Directory

[SAFE]: The skill is a workflow orchestration tool for GitHub repositories. It follows standard development practices such as fetching the latest changes (git pull), validating branch naming conventions, and utilizing existing repository templates for issues and PRs.
[COMMAND_EXECUTION]: The skill executes standard git and GitHub CLI (gh) commands to manage the development lifecycle. These include git pull, git branch, gh auth status, and gh repo view. These are routine operations for an AI agent tasked with repository management.
[DATA_EXFILTRATION]: The skill includes explicit safety guidelines prohibiting the inclusion of sensitive data like .env files, API keys, or personal information in issues, commits, or PRs. It limits remote write operations to those explicitly requested by the user.
[PROMPT_INJECTION]: No malicious prompt injection patterns were detected. The instructions focus on structured branching and planning logic without attempting to bypass safety filters or override global system instructions.
[SAFE_PRACTICE]: The skill demonstrates safe practices by requiring user confirmation before creating remote resources (like issues) when the working directory is clean, and by avoiding direct commits to protected main branches.

global-branch-planning-pipeline