The Agent Skills Directory

[DATA_EXPOSURE]: The skill facilitates the reading of local image files for processing. This capability is secured by a validation mechanism in core/image.py that resolves paths and enforces a blacklist of sensitive system directories (e.g., /etc, /proc, /sys) and user configuration directories (e.g., .ssh, .aws, .gnupg). It also explicitly blocks access to .env files.
[DYNAMIC_EXECUTION]: The model registration system utilizes dynamic Python imports. This is implemented safely by restricting the imports to a predefined, hardcoded list of internal skill files, preventing the execution of arbitrary or untrusted code.
[COMMAND_EXECUTION]: The skill utilizes the uv package manager for isolated dependency management and execution of its internal scripts. This follows modern Python development best practices.
[CREDENTIALS_UNSAFE]: No hardcoded credentials were found. The skill correctly instructs users to manage API keys via environment variables or .env files, which is the standard safe practice for such integrations.

mulerouter