Proto-OKN MCP Server

A single MCP server that lets Claude discover, inspect, and query 30+ NSF-funded Prototype Open Knowledge Network (Proto-OKN) scientific knowledge graphs — including SPOKE biomedicine and BioBricks ICE chemical safety — in natural language.

   
Type MCP server (remote, streamable HTTP; self-host also available)
Supplier SBL-SDSC — Structural Bioinformatics Laboratory, San Diego Supercomputer Center
Availability Beta — hosted endpoint live; source repo sbl-sdsc/mcp-proto-okn
Pricing Free / OSS (BSD 3-Clause); hosted endpoint is free but requires a Claude Pro/Max subscription for custom connectors
Capabilities Read-only — knowledge-graph discovery, SPARQL query, cross-graph identifier bridging
Verified works · 2026-07-20
Security cleared · 2026-07-20 — provenance matches SDSC, BSD-3, maintained, read-only over public KGs, no OSV advisories

How to install

  • Claude.ai / Claude Desktop — Settings → Connectors → Add custom connector. Name it proto-okn, server URL https://apps.okn.us/okn-mcp/mcp, then click Configure and set tool permissions to Always allow. In a new chat, toggle proto-okn on. (Requires a Claude Pro or Max subscription for MCP connectors.)
  • Claude Code — direct remote MCP add: 
    claude mcp add --transport http proto-okn https://apps.okn.us/okn-mcp/mcp

    This is a long-lived hosted service — nothing to run locally. After adding, run /mcp to confirm the server is connected. (Equivalently, add a "proto-okn" entry with "type": "url" and "url": "https://apps.okn.us/okn-mcp/mcp" under mcpServers in your project .mcp.json.)

  • Self-host (optional) — clone sbl-sdsc/mcp-proto-okn and follow docs/develop.md to run the server locally; only needed if you want your own instance rather than the hosted endpoint above.

What it does

Exposes 30+ Proto-OKN knowledge graphs (hosted on the FRINK federation platform, catalogued in the OKN Knowledge Graph Registry) through one unified interface. Capabilities:

  • Graph discovery and filtering — list and route to relevant knowledge graphs
  • Schema inspection — inspect a graph’s classes, predicates, and structure
  • SPARQL execution — run queries against a chosen graph
  • Cross-graph identifier bridging — link IDs across graphs
  • Multi-graph querying — combine results from several sources
  • Ontology-driven query expansion — automatically expand queries via MONDO, HP, GO, UBERON, and ChEBI hierarchies (through Ubergraph) so a search for a parent concept matches all descendants
  • Transcript / schema-visualization generation

Covered graphs include SPOKE (Scalable Precision Medicine Open Knowledge Engine), BioBricks ICE (chemical safety / cheminformatics), DREAM-KG, and SAWGraph (agricultural products and water monitoring).

Primary use cases: Cross-domain scientific knowledge-graph Q&A, SPARQL query authoring, biomedical/chemical-safety knowledge integration.

Notes

Beta as of 2026-07-18; the team welcomes feedback and bug reports via GitHub issues. The hosted endpoint (https://apps.okn.us/okn-mcp/mcp) requires no local install and no API key, but visiting it in a browser returns a JSON-RPC / HTTP error (it speaks the MCP protocol, not plain HTTP GET) — that is expected and does not indicate an outage. The MCP URL and setup steps have changed over the project’s life; check the GitHub repo for the latest.

Surfaced from user request #50 (@goodb). The request suggested the General-Purpose Utilities shelf, but the server is a cross-domain life-science knowledge-graph gateway (SPOKE biomedicine, BioBricks chemical safety, ChEBI/MONDO/GO/HP/UBERON ontology expansion), so it is tagged All.

Part of the NSF Proto-OKN program. Described in arXiv preprint 2605.30283.

Sources

Installed this tool?

Share feedback — install path, OS, errors, workarounds. The form opens with this tool pre-selected and a link back to this page.