The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes the kicad-cli utility via subprocess.run in sch.py. This execution is limited to specific subcommands (sch erc and sch export svg) required for the skill's primary functionality of validating schematics and performing linting. The tool path is determined by checking standard KiCad installation locations or a user-defined KICAD_CLI environment variable.\n- [INDIRECT_PROMPT_INJECTION]: The skill has a surface for indirect prompt injection as it processes KiCad schematic files (.kicad_sch) which could originate from untrusted sources.\n
Ingestion points: Schematic initialization in sch.py reads files from the project directory.\n
Boundary markers: No explicit markers are used as the skill treats the schematic content as structured data rather than instructions.\n
Capability inventory: Includes file writing (overwriting the schematic) and execution of the local kicad-cli binary.\n
Sanitization: Employs a custom S-expression parser and formatter that properly escapes string literals (e.g., in the q function), which mitigates risks associated with malicious data corrupting the file structure or triggering unintended behavior in the CAD software.

kicad-schematic