Connect AI to SciSure Research

SciSure MCP: Overview · ChatGPT (OpenAI) · Claude · LM Studio · Open WebUI · Other MCP applications

Connect the AI tools your lab already uses to SciSure Research through the Model Context Protocol (MCP). Each guide walks through the connection, authentication, and first successful request for one application.

📘

Private Cloud and On-Premises only

MCP — and the OAuth 2.1 integration it relies on — is available on Private Cloud and On-Premises SciSure Research installations, not on the shared Cloud environment.

Choose your application

Commercial AI

• ChatGPT (OpenAI) — add SciSure Research as a custom MCP plugin and sign in with a SciSure OAuth client.
• Claude — add SciSure Research as a custom connector and authorize it against your instance.

Local AI

• LM Studio — add SciSure to mcp.json, authenticate the desktop app, and call research tools locally.
• Open WebUI — configure SciSure Research as an external MCP tool server.

Other MCP applications

• Another MCP-compatible application — generic connection details for any client that supports remote MCP servers with OAuth.

How the connection works

1. Your AI application connects to your instance's MCP endpoint: https://<your-instance>/api/v1/mcp.
2. It signs the user in through a SciSure OAuth client that a group administrator registers in SciSure Research.
3. The application then calls the research tools the user is authorized for — experiments, protocols, samples, and tasks.

The OAuth 2.1 developer guide covers the authorization flow, tokens, and scopes in depth.