The Agent Skills Directory

[SAFE]: The skill is designed with safety as a primary feature, specifically Step 2 which instructs the agent to scan all changes for secrets, credentials, and personal data before proceeding. It requires user interaction if such data is found.
[COMMAND_EXECUTION]: The skill utilizes standard Git commands (git status, git diff, git commit, git push) and an external utility prune-tests. These operations are directly aligned with the skill's description and intended functionality as a deployment helper.
[INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data in the form of local code changes and Git diffs. While this is an ingestion surface, the skill's logic focuses on data inspection and sensitive content detection rather than interpreting the code as instructions, and it enforces human-in-the-loop confirmations for destructive or sensitive actions.

ship