kicad-tools
The kct CLI and a Python library for agents: parse and edit KiCad schematics and PCBs with JSON output, run DRC and ERC without a running KiCad
Medium risk
We rate an entry medium when the tool runs code, makes network calls or reads project files. Check what exactly it does before installing.
Why this level
- Runs code and reads and writes project files
- The datasheet features reach the network
Install
In your terminal, with SkillFoxx CLI
npx skillfoxx add cli/kicad-toolsDetects the agents on your machine, checks the risk and pins the version.
Other ways to install
Install the tool
pipx install kicad-toolsWithout pipx, pip install kicad-tools works too.
Install the CLI with pip install kicad-tools, then use the kct command, for example kct drc board.kicad_pcb --mfr jlcpcb --format json. For the MCP server run pip install "kicad-tools[mcp]" and start it with kct mcp serve.
Other ways from the author
pip install kicad-toolsBase install. Optional extras add features: kicad-tools[native], [placement], [mcp], [all].
This is third-party code. Review the repository files before installing.
What it does
The toolkit parses KiCad files into clean Python objects and edits schematics and boards without a GUI, and every CLI command can emit JSON. Through the kct command you get symbol listings, net tracing, a bill of materials, ERC, and DRC against a chosen manufacturer's rules. It adds an autorouter, placement optimization by physics-based and evolutionary methods, trace optimization, ready circuit blocks, and IPC-7351 footprint generators. A separate module exposes a reasoning interface where an LLM makes placement decisions while the tools handle geometry. It also ships an MCP server and datasheet tooling for search, download, and PDF parsing.
Who it is for. For hardware engineers and developers who automate KiCad work or build agents that design PCBs.
Good fit when
- You need to parse or modify a KiCad schematic or board from a script and get JSON
- You need to run ERC or DRC against manufacturer rules in CI
- An agent should reason about board placement and routing
Not a fit when
- You need manual work in the GUI editor rather than automation
- The project is unrelated to KiCad and electronics
Example request
Run DRC on board.kicad_pcb against jlcpcb rules and return the violations as JSONLimitations
ERC and some checks rely on the external kicad-cli; without it they are skipped. GPU acceleration, the native C++ router, placement optimization, and the MCP server install as separate extras. The datasheet features reach the network for search and download. The project is marked beta.
How to disable. Remove the package with pip uninstall kicad-tools. If you configured the MCP server, delete the kicad-tools entry from your agent's config.
Security check
- Runs code and reads and writes project files
- The datasheet features reach the network
README in short
The README presents standalone Python tools for agents that work with KiCad schematics and boards and emit machine-readable JSON without a running KiCad. It covers the kct command, the Python API, the autorouter, an LLM reasoning module, an MCP server, circuit blocks, whole-project workflow, footprint generators, pure-Python DRC, placement and trace optimization, and datasheet tooling. The package installs through pip with optional extras for GPU, the native router, and placement. It is MIT licensed, needs Python 3.10 or newer, and has a gallery of example boards on a separate site.
FAQ
Do I need KiCad installed?
Parsing, editing, and pure-Python DRC work without a running KiCad. ERC and some checks need the external kicad-cli.
Can I use it as an MCP server?
Yes, install the mcp extra and run kct mcp serve or python -m kicad_tools.mcp.server, then add the server to your agent config.
Related
A skills library that gives coding agents a development process: brainstorming, planning, TDD, subagents and code review
Skills for real engineers by Matt Pocock
Skills For Real Engineers
Small composable skills for engineering with agents: plan grilling, TDD, bug diagnosis, code review and architecture
GitHub toolkit for spec-driven development: the specify CLI adds agent commands and skills to a project, from principles to implementation
Reference MCP servers
Model Context Protocol servers
Official reference MCP servers: Filesystem, Fetch, Git, Memory, Sequential Thinking, Time and Everything