The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses subprocess.run to orchestrate various stages of the documentation pipeline.
scripts/kidoc_scaffold.py runs analysis scripts on schematic and PCB files.
scripts/kidoc_generate.py dispatches document rendering to format-specific scripts (PDF, DOCX, etc.).
scripts/kicad_cli.py invokes the system kicad-cli tool.
[EXTERNAL_DOWNLOADS]: scripts/kidoc_venv.py automatically sets up a project-local virtual environment and installs required packages from the official Python Package Index (PyPI). This is a standard and safe practice for managing dependencies.
[PROMPT_INJECTION]: The skill assembles data from KiCad project files into context packages used by the AI agent to write design narratives. This creates a surface for indirect prompt injection where malicious design data could influence the agent.
Ingestion points: scripts/kidoc_narrative.py reads component descriptions, net names, and analysis findings from KiCad hardware designs.
Boundary markers: Absent. User-controlled design data is directly interpolated into the narrative context without clear delimiters or 'ignore embedded instructions' warnings.
Capability inventory: The skill possesses extensive file system access and the ability to execute subprocesses through orchestration scripts like scripts/kidoc_generate.py and scripts/kidoc_scaffold.py.
Sanitization: Hardware design data is not sanitized or escaped for potential instructions before being passed to the AI agent as narrative context.

kidoc