Pinscope
A comprehensive datasheet-backed electrical schematic reviewer
What is it?
What it is
An automated schematic review tool that validates circuit designs against manufacturer datasheets.
Why it exists
Standard Electrical Rule Checks (ERC) in EDA tools cannot detect datasheet-specific violations, such as voltage mismatches between ICs, incorrect pin alternate-functions, or missing bypass capacitors, which often require manually reading hundreds of pages of documentation.
Who should use it
Hardware EngineersElectrical EngineersPCB DesignersElectronics Hobbyists
Who should avoid it
Users without an Anthropic API keyUsers unable to provide netlists or BOMsThose seeking a 100% automated 'oracle' without manual verification
How it works
A quick walkthrough in plain English
How Pinscope works
Step 1 of 3
You interact with it
Open Pinscope, send a request, or connect it to your stack.
Features
Advantages
- Detects non‑rule violations that EDA ERC tools miss (e.g., voltage mismatches, alternate‑function mismatches, unconnected bypass pins)
- Citations to datasheet pages give engineers confidence and traceability
- Deterministic calculations avoid model‑based uncertainty
- Supports a wide range of EDA tools and file formats
- Local execution keeps data private; hosted version offers collaboration
- Open‑source code encourages community contributions and transparency
- Caching of extracted data improves performance over time
- Clear separation of review logic and data extraction
- Provides a deterministic BOM roll‑up useful for cost and supply chain analysis
Disadvantages
- Requires an Anthropic API key and a paid plan for large PDF extraction
- Needs Python 3.12+ and Node 20+, which may limit older environments
- Setup involves multiple steps (pip install, env file, uvicorn, npm install) that can be complex for novices
- AGPL license may restrict commercial use without a commercial license
- Extraction accuracy depends on the quality of the PDF and the prompts; may miss or misinterpret data
- Performance can be slower for very large designs due to PDF parsing and graph traversal
- Findings are not guaranteed; the tool may miss issues or flag false positives
- Requires manual provision of datasheet PDFs or DigiKey API keys for automatic retrieval
- Limited to IC review; passive components and other design aspects are not automatically checked
Installation
cloud
pinscope.ai
docker
not found
native
pip install -r backend/requirements.txt cp backend/.env.example.env python3 scripts/upload_skills.py --update python3 -m uvicorn backend.main:app --reload cd frontend && npm install && npm run dev
compose
not found
FAQ
What files do I need to run Pinscope on my design?
You need a BOM file (CSV or XLSX) and a netlist file exported from your EDA tool (PADS-PCB .asc, EDIF 2.0.0 .edn, KiCad, Altium, OrCAD, Allegro, Xpedition, EasyEDA, Eagle). Additionally, you must provide the datasheet PDFs for each IC in the design, or set up optional DigiKey API keys so Pinscope can fetch them automatically.
How does Pinscope parse my netlist and BOM?
Pinscope reads the BOM and netlist into a bipartite graph of components and nets. It supports common formats such as PADS-PCB .asc, EDIF 2.0.0 .edn, and exports from KiCad, Altium, OrCAD, Allegro, Xpedition, EasyEDA, and Eagle. The graph allows querying connected components, nets for pins, and pin tables during the review.
What is the process for extracting datasheet information?
Pinscope extracts pin tables and specifications from the PDF datasheets. It first trims large datasheets to the relevant pages, then parses the content. Extracted data is cached in a shared library so each part number is processed only once, speeding up subsequent reviews.
Can I use a hosted version of Pinscope?
Yes. The hosted service at https://pinscope.ai runs the same code with team workspaces and a pre‑warm shared parts library. It removes the need to install and run the local stack, but requires an account and may involve billing.
What software versions and dependencies are required?
You need Python 3.12+ and Node 20+ to run the backend and frontend. Install the Python dependencies with `pip install -r backend/requirements.txt`, set your Anthropic API key, and start the server with `uvicorn backend.main:app --reload`. The frontend runs with `npm install && npm run dev`.
How can I add new parts or update the extraction library?
Run `python3 scripts/upload_skills.py --update` once to register extraction prompts under your account. The extraction library is cached in the `data/` directory; adding new PDFs or updating existing ones will automatically refresh the cache on the next review.
Featured in Videos
YouTube tutorials and walkthroughs for Pinscope
Alternatives
Similar projects ranked by category, topics, and text overlap.
