The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes user-supplied Studio 5000 Logix Designer project files in XML format (.L5X).
Ingestion points: The inspect_l5x_project function in gemini.md parses external .L5X files using xml.etree.ElementTree to audit tags, UDTs, and Add-On Instructions.
Boundary markers: No explicit delimiters or instructions are provided to the agent to distinguish between project data and potential malicious instructions embedded in XML comments or tag names.
Capability inventory: The skill can execute local industrial software (LogixDesigner.exe), establish network connections via Python's socket module, and perform read/write operations on industrial controllers via pycomm3.
Sanitization: The skill performs standard XML parsing but lacks specific sanitization or filtering to prevent the agent from interpreting natural language instructions embedded within the industrial project metadata.
[COMMAND_EXECUTION]: The skill utilizes command-line instructions to interface with industrial hardware and software.
Evidence: SKILL.md provides commands to launch LogixDesigner.exe with specific project paths and uses a Python one-liner to check connectivity on TCP port 44818 (CIP protocol).
Context: These operations are restricted to the intended industrial automation use case and involve local environment management. A heuristic detection for a reverse shell was evaluated and confirmed to be a false positive triggered by the standard socket.connect port-knocking check.

rockwell-studio-5000