The Agent Skills Directory

[COMMAND_EXECUTION]: The skill employs pcb_execute_tools to run dynamically generated JavaScript code. This is the primary method for interacting with the KiCad API but constitutes a runtime code execution surface.
[COMMAND_EXECUTION]: The skill exhibits an indirect prompt injection surface by ingesting data from KiCad board files (via kicad_get_board_summary and kicad_query in SKILL.md) without explicit boundary markers or sanitization. This data could potentially contain malicious instructions that influence the agent's use of modification tools like kicad_update_items.
[SAFE]: No evidence of data exfiltration, credential exposure, or unauthorized network communication was found. The skill's functionality is limited to the defined MCP toolset.

kicad-layout