The Agent Skills Directory

[COMMAND_EXECUTION]: The skill generates a local shell script and uses chmod +x to make it executable. This script contains logic to parse tool calls and block specific Git patterns.
[COMMAND_EXECUTION]: It uses the jq utility to modify the agent's configuration files (settings.json), registering the script as a persistent hook for the Bash tool.
[SAFE]: No network activity or external data exfiltration was detected. All operations are local to the user's filesystem and targeted at the intended application's configuration directories.
[SAFE]: The behavior is transparent, documented, and aligned with the stated goal of providing safety guardrails.

git-guardrails