kicad-schematic
Warn
Audited by Gen Agent Trust Hub on Feb 26, 2026
Risk Level: MEDIUMCOMMAND_EXECUTIONREMOTE_CODE_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill instructs the agent to execute shell commands with administrative privileges for environment setup.\n
- Evidence: The SKILL.md file contains instructions for macOS and Linux to use sudo ln -sf to make kicad-cli available on the system PATH.\n
- Evidence: Setup scripts include path detection commands like which and where to locate existing installations.\n- [COMMAND_EXECUTION]: The helper library executes external system commands via subprocess calls to perform validation tasks.\n
- Evidence: The run_erc function in scripts/kicad_sch_helpers.py uses subprocess.run to call the kicad-cli tool with parameters derived from project files.\n- [REMOTE_CODE_EXECUTION]: The skill's primary workflow relies on the agent writing and executing local Python scripts to generate schematic content.\n
- Evidence: SKILL.md and README.md detail a process where the agent generates Python code using the SchematicBuilder class and runs it locally to create .kicad_sch files.\n- [PROMPT_INJECTION]: The skill contains instructions designed to enforce specific behaviors using strong directives that could interfere with agent safety constraints.\n
- Evidence: Use of markers like Critical Principle and NON-NEGOTIABLE in SKILL.md to dictate agent behavior during coordinate transformations and library parsing.\n- [PROMPT_INJECTION]: The skill possesses an attack surface for indirect prompt injection through the processing of external KiCad symbol files.\n
- Ingestion points: The SymbolLibrary.load_from_kicad_sym method in scripts/kicad_sch_helpers.py reads data from .kicad_sym files.\n
- Boundary markers: No delimiters or instructions to ignore embedded content are present in the file processing logic.\n
- Capability inventory: The skill has the ability to execute system commands via subprocess.run and supports local script generation and execution.\n
- Sanitization: Content from the library files is used to drive script generation and schematic construction without explicit sanitization.
Audit Metadata