The Agent Skills Directory

[SAFE]: The skill incorporates defensive security protocols, such as the gitignore-audit.ts utility and secret exposure audits described in exposure-audit.md, which proactively prevent the leakage of credentials and sensitive configuration files.
[COMMAND_EXECUTION]: Automated Git and GitHub CLI operations are used to manage the development lifecycle, including branch creation, PR management, and production releases, all within expected usage patterns for a developer tool.
[INDIRECT_PROMPT_INJECTION]: The skill manages potentially untrusted data from GitHub issues and pull request threads. Ingestion points: The skill uses gh issue view and gh pr view to fetch external content as described in anti-slop-triage.md. Boundary markers: It implements a structured 9-tier 'Anti-Slop' triage matrix to classify data before execution. Capability inventory: Access to bash, run_command, view_file, and write_to_file tools. Sanitization: Mandatory classification of issues into categories like 'spam', 'generated-slop', or 'unsafe-or-security-sensitive' provides a robust filter against injection attempts.

git