The Agent Skills Directory

[COMMAND_EXECUTION] (SAFE): The Python scripts package_skill.py and quick_validate.py perform standard file system operations and data validation. There is no evidence of arbitrary shell command execution or subprocess hijacking.- [DATA_EXFILTRATION] (SAFE): No network operations (curl, wget, requests) or access to sensitive file paths (e.g., ~/.ssh, credentials) were found. Data processing is confined to local skill directories.- [REMOTE_CODE_EXECUTION] (SAFE): The scripts do not download or execute remote code. Dependencies are limited to standard libraries and PyYAML, which is used safely via yaml.safe_load().- [PROMPT_INJECTION] (SAFE): The documentation files (output-patterns.md, workflows.md) contain educational examples for skill behavior and do not include instructions aimed at bypassing agent safety filters or overriding system prompts.

skill-creator