Metadata-Version: 2.4
Name: medxpert-reg-hub
Version: 2.0.2
Summary: Medical-device regulatory MCP server: NMPA UDI, FDA 510(k)/PMA/classification, 20-hub knowledge-base lookups (read-only).
License: MIT
Project-URL: Homepage, https://medxpert.cn
Project-URL: Repository, https://github.com/stevenzhao07/medxpert-reg-hub
Keywords: mcp,mcp-server,medical-device,regulatory,nmpa,fda,510k,pma,udi
Classifier: Development Status :: 4 - Beta
Classifier: Environment :: Console
Classifier: Intended Audience :: Healthcare Industry
Classifier: License :: OSI Approved :: MIT License
Classifier: Programming Language :: Python :: 3
Classifier: Topic :: Scientific/Engineering :: Medical Science Apps.
Requires-Python: >=3.10
Description-Content-Type: text/markdown
Requires-Dist: mcp>=2.1.1

# medxpert-reg-hub · MCP Server

`mcp-name: io.github.stevenzhao07/medxpert-reg-hub`

Read-only medical device regulatory MCP server. Tools (all read-only):

- `udi_lookup` — NMPA UDI queries (main DI / product name / registration no.)
- `fda_510k_lookup` — FDA 510(k) queries (K-number / company / product)
- `pma_lookup` — FDA PMA queries (PMA number / company / product)
- `class_lookup` — FDA product classification queries (product code / name)
- `hub_search` — keyword search across the 20-hub regulatory knowledge base (CN/US/EU/JP)

## Install & run

```bash
pip install medxpert-reg-hub
medxpert-mcp        # stdio MCP server
```

## Data (MEDXPERT_DB / MEDXPERT_REFS)

The PyPI wheel is metadata-light by design (the full dataset is ~600M+ rows across
SQLite). Point the server at your data:

```bash
export MEDXPERT_DB=/path/to/medxpert_device.sqlite
export MEDXPERT_REFS=/path/to/skills/references
medxpert-mcp
```

The reference data bundle is published in the GitHub repo
[stevenzhao07/medxpert-reg-hub](https://github.com/stevenzhao07/medxpert-reg-hub)
(docs/ and skills/references/), so you can clone it and set the env vars, or use
your own RA knowledge base.

License: MIT © 2026 MedXpert
