kicad
Pass
Audited by Gen Agent Trust Hub on Mar 14, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: Indirect Prompt Injection Surface (Category 8). The skill processes untrusted external data from KiCad design files and PDF documentation, which could be used to influence agent behavior.
- Ingestion points: Analysis scripts (analyze_schematic.py, analyze_pcb.py, analyze_gerbers.py) extract metadata, labels, and descriptions from user-provided KiCad and Gerber files. The agent also reads content from PDF schematics.
- Boundary markers: The analyzer outputs structured JSON, but the instructions do not require the agent to treat field content (such as component descriptions or net names) as untrusted data or use specific delimiters to prevent command interpretation.
- Capability inventory: The skill provides Python scripts for file analysis and encourages the agent to use parsed data to interact with other skills (e.g., DigiKey, BOM).
- Sanitization: No specific sanitization is performed on text strings extracted from analyzed files before they are presented to the agent.
Audit Metadata