The Agent Skills Directory

[COMMAND_EXECUTION]: The script scripts/trigger-test.py executes external CLI tools (claude, codex, copilot) using subprocess.run(). This is a core part of its functionality for testing prompt triggers and follows safe practices by using list-based argument passing instead of shell invocation.
[SAFE]: The skill's primary content consists of instructional documentation that promotes safe and predictable agent behavior through concepts like completion criteria and progressive disclosure. The provided Python script is a standard developer tool for validation and does not perform any unauthorized network or file operations.

agent-docs