The Agent Skills Directory

[REMOTE_CODE_EXECUTION]: The scripts/gen_urdf/cli.py script utilizes importlib to dynamically load and execute code from arbitrary Python files. Specifically, it calls module_spec.loader.exec_module(module) to execute the gen_urdf() function in the target file. This pattern allows for arbitrary code execution within the agent's environment. \n- [COMMAND_EXECUTION]: The skill provides a command-line interface (scripts/gen_urdf) that the agent is instructed to use. This command interacts with the filesystem and executes Python logic that can be influenced by external file content. \n- [PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection (Category 8). If an attacker influences the content of a Python or URDF file that the agent processes, they can achieve unauthorized code execution. The XML parsing in scripts/urdf_source.py uses xml.etree.ElementTree, which lacks protection against XML External Entity (XXE) attacks. \n
Ingestion points: scripts/gen_urdf/cli.py (Python files) and scripts/urdf_source.py (URDF files). \n
Boundary markers: None are used to separate untrusted file content from instructions. \n
Capability inventory: Arbitrary Python execution (via importlib) and filesystem write operations. \n
Sanitization: The skill performs basic path and suffix checks but does not sanitize the content of the files it executes or parses.

urdf