The Agent Skills Directory

[COMMAND_EXECUTION]: The diagnostic script scripts/check_env.py utilizes subprocess.run to verify Python version information and binary paths using which or where. These calls are restricted to a safe, hardcoded list of diagnostic commands.\n- [SAFE]: The skill correctly handles sensitive API tokens by storing them in a local YAML configuration file (~/.openclaw/bt-skills.yaml) rather than hardcoding them in scripts.\n- [SAFE]: All network operations are directed to server hosts explicitly configured by the user via the bt-config.py utility. There are no unauthorized external connections or data exfiltration patterns.\n- [SAFE]: The use of __import__ in environment checks is restricted to verifying that required dependencies like requests and pyyaml are installed.\n- [SAFE]: The skill instructions in SKILL.md include appropriate safety guidelines for the AI, emphasizing data neutrality and user notification before command execution.

btpanel