The Agent Skills Directory

[SAFE]: The skill proactively warns against committing sensitive files like .env and credentials.json to prevent accidental credential leakage.- [SAFE]: It instructs the agent to stage files individually by name rather than using bulk commands, reducing the risk of including unauthorized or sensitive content.- [SAFE]: The commit creation logic uses a quoted HEREDOC (cat <<'EOF'), which is a secure practice to ensure the commit message content is not interpreted by the shell.- [SAFE]: While the skill ingests data from repository files via git diff, it follows a structured workflow for analysis and uses safe command patterns for subsequent actions.

commit