The Agent Skills Directory

[COMMAND_EXECUTION] (SAFE): The skill contains example commands such as brew install, sudo apt install, and ln -s for symlinking. These are identified as benign documentation examples using placeholder names like tool-name or my-tool. They are intended to show the user or agent how to structure installation and usage instructions in a new skill.
[CREDENTIALS_UNSAFE] (SAFE): The references/template.md file contains a placeholder for an environment variable (TOOL_API_KEY="your-key"). This is a standard documentation practice and does not involve hardcoded secrets.
[DATA_EXFILTRATION] (SAFE): While the documentation mentions accessing config directories (~/.config), there are no patterns indicating the unauthorized transmission of data to external domains.
[PROMPT_INJECTION] (SAFE): No instructions designed to override the agent's safety guardrails or system instructions were found. The instructional language used is appropriate for a tutorial or guide.

writing-cli-skills