Check a circuit board design against every chip's datasheet before manufacturing it.
Catch voltage mismatches between connected chips that standard rule checkers miss.
Compute a capacitor voltage derating table automatically from the design graph.
Get page-cited findings so an engineer can verify each flagged issue against the source datasheet.
| faradworks/pinscope | colossus-lab/openarg_backend | mr-un1k0d3r/azureredops | |
|---|---|---|---|
| Stars | 135 | 135 | 135 |
| Language | Python | Python | Python |
| Setup difficulty | moderate | hard | moderate |
| Complexity | 4/5 | 5/5 | 4/5 |
| Audience | developer | developer | ops devops |
Figures from each repo's GitHub metadata at analysis time.
Requires Python 3.12+, Node 20+, and an Anthropic API key to run the review pipeline.
Pinscope reviews electronic circuit schematics the way an experienced hardware engineer would, by actually reading the datasheets for every part on the board instead of only checking general wiring rules. You give it a netlist and a bill of materials describing your design, along with the datasheet PDFs for the chips involved, and it builds a graph of how every component and wire connects to every other one. The tool exists because standard electrical rule checks in circuit design software can pass a board that still does not work. A checker has no way of knowing that a specific chip drives its output at a voltage too high for the microcontroller pin it connects to, or that a signal has been wired to a pin whose datasheet only allows a different function, or that leaving a certain pin unconnected will quietly add a large amount of electrical noise. None of that breaks a formal wiring rule, but it is written in the part's datasheet, and nobody has time to reread hundreds of pages of documentation for every part on every design revision. Pinscope works in four steps. First it parses the bill of materials and netlist, which can come from tools like KiCad, Altium, OrCAD, or EasyEDA, into a graph of components and connections. Second, it pulls the relevant pin tables and specifications out of each datasheet PDF, caching the result so a given part is only processed once. Third, it reviews each chip individually, looking at its datasheet alongside the surrounding circuit and filing findings with a severity level, its reasoning, and the exact datasheet page that supports it. Fourth, it computes things that do not need judgment at all, like a full parts list summary and a capacitor voltage derating table, directly from the graph with no AI model involved. A final pass can merge duplicate findings or lower their severity, but never raise it. Running it locally requires Python 3.12 or newer, Node 20 or newer, and an Anthropic API key, since the per-chip review step uses Claude. A hosted version with team accounts is available at pinscope.ai. The project itself is released under AGPL-3.0, with separate commercial licensing available by contacting the developer.
An AI-assisted schematic reviewer that reads component datasheets to catch wiring problems standard electrical rule checks miss.
Mainly Python. The stack also includes Python, FastAPI, React.
Free to use and modify, but if you run a modified version as a network service you must share your source code too, commercial licensing is available separately.
Setup difficulty is rated moderate, with roughly 30min to a first successful run.
Mainly developer.
This repo across BitVibe Labs
Don't trust strangers blindly. Verify against the repo.