The Agent Skills Directory

[SAFE]: The skill is strictly limited to an advisory role. It contains explicit instructions to never execute tasks, create files, or run commands, ensuring it remains within its intended scope of prompt optimization.\n- [COMMAND_EXECUTION]: The skill performs automated project detection by reading common configuration files such as package.json, go.mod, and requirements.txt to identify the technology stack. These are standard development environment files and do not contain sensitive user credentials or private keys.\n- [PROMPT_INJECTION]: Instructions are provided to handle cases where a user might attempt to force the agent into implementation mode, directing the agent to remain in its advisory capacity and refuse direct execution requests.

prompt-optimizer