AI-agent skills and scripts for searching the JLCPCB / LCSC parts catalogue.
Point your coding agent at a BOM (KiCad, CSV, schematic netlist, etc.) and ask it to resolve every line to a part that is currently in stock at JLCPCB — preferring Basic parts to avoid extended-part fees, matching package, value, tolerance, voltage, and other parametric constraints.
Example prompt:
Fill the BOM with parts available from JLCPCB.
Prefer Basic parts. For each line, output MPN, LCSC number, stock, and Basic/Extended.
Find 0.1uF 0603 X7R 25V+ Basic capacitor on JLCPCB.
Ask your coding agent to update the local SQLite catalogue from the upstream JSON shards.
Update the local JLCPCB parts catalogue.
In Claude Code:
/plugin marketplace add mash/jlcpcb-skills
/plugin install jlcpcb-skills@jlcpcb-skills
Available next session. The skill installs the jlcpcb-parts CLI itself on first use, picking the method (pipx / uv / venv pip / …) that fits your Python environment. To skip bootstrap, install git+https://github.com/mash/jlcpcb-skills.git yourself however you like.
The local SQLite is built from the JSON shards published by yaqwsx/jlcparts at https://yaqwsx.github.io/jlcparts/data/ (refreshed every ~6h upstream). yaqwsx/jlcparts is licensed under the MIT License (Copyright (c) 2024 Jan Mrázek).
This repository does not redistribute the catalogue data — each user fetches it directly from the upstream CDN on first use of the update command.
MIT — see LICENSE.