The Agent Skills Directory

[SAFE]: The skill manages internal state files ('ring--llm.md') within the project's own directory structure. It uses absolute paths for system tools like /usr/bin/find and /usr/bin/grep to ensure predictable behavior.
[SAFE]: The execution of python3 ${CLAUDE_PLUGIN_ROOT}/scripts/ring.py refers to a script bundled with the plugin itself, which is a standard method for performing validation and scanning tasks within a specialized toolset.
[SAFE]: No external network operations, credential exposures, or obfuscation techniques were identified. The instructions focus on application-specific logic such as write tokens and state transitions (e.g., 'inline' to 'resumed').

sync