Koskamo

BGPT MCP

Search scientific papers with structured experimental data from full-text studies. Retrieve claims linked to experiments, exact results, scope, limitations, falsification criteria, and source provenance to build verifiable scientific AI.

Endpoint and access

No sign-in to connect and list tools. Individual tools may still need a vendor account, credentials, or permissions; check the documented prerequisites.

BGPT MCP logo, by BGPTCommunity listingLegacy SSENo sign-inEndpoint checked 2026-11-01Published by BGPT
MCP endpoint (Legacy SSE)
https://bgpt.pro/mcp/sse
Vendor documentation

What you can do with it

  • Search by topic or DOI
  • Retrieve claims linked to experiments
  • Trace evidence to source sections
  • Build scientific AI with verifiable evidence

Source and maintenance

Community listingRegistry 1.1.0
Maintenance checked

Protocol and capabilities

Server
BGPT 2.14.7
MCP protocol
2025-06-18
Tools
2 advertised
Resources
0 advertised
Resource templates
0 advertised
Prompts
0 advertised
Server-advertised capabilities
{
  "tasks": {
    "list": {},
    "cancel": {},
    "requests": {
      "tools": {
        "call": {}
      },
      "prompts": {
        "get": {}
      },
      "resources": {
        "read": {}
      }
    }
  },
  "tools": {
    "listChanged": true
  },
  "prompts": {
    "listChanged": true
  },
  "resources": {
    "subscribe": false,
    "listChanged": true
  },
  "experimental": {}
}

Server-provided instructions

Untrusted server output shown for inspection. It is not Koskamo guidance and is excluded from knowledge search.

BGPT is the evidence layer for scientific AI. It turns full-text papers into versioned records linking claims to experiments, exact results, scope, limitations, falsification criteria, and source provenance. Use output_format='evidence' for compact records for questions, reviews, retrieval, evaluation, and permitted training. Search records are returned in results; DOI lookup records are returned in result. 50 free results per network, then $0.02 per result with an API key. Get an API key at https://bgpt.pro/mcp

Tools

2 advertised

search_papers — Search scientific evidence
Advertised by the server during the last successful anonymous check.
Advertised read-only

Search claim-level evidence extracted from full-text scientific papers. Args: query: Search terms (e.g. "CRISPR gene editing efficiency"). SHORT, concise queries are best. English language only. Use days_back, num_results, min_citations, and study_type instead of adding years or filters to the query. num_results: Number of results to return (1-100, default 16). First 50 results are free, then metered per result for paid users. days_back: Only return papers published within the last N days. min_citations: Only return papers with at least this many references cited. study_type: Only return papers of this study type. One of: primary study | systematic review | meta-analysis | narrative review | protocol | dataset | commentary | other. output_format: "evidence" for compact claim-level evidence (default), "legacy" for original paper metadata, or "full" for both. Returns: An envelope whose results list contains papers with claims, experiments, exact results, demonstrated scope, limitations, and provenance.

search_papers input schema
{
  "type": "object",
  "required": [
    "query"
  ],
  "properties": {
    "query": {
      "type": "string"
    },
    "days_back": {
      "anyOf": [
        {
          "type": "integer"
        },
        {
          "type": "null"
        }
      ],
      "default": null
    },
    "study_type": {
      "anyOf": [
        {
          "type": "string"
        },
        {
          "type": "null"
        }
      ],
      "default": null
    },
    "num_results": {
      "type": "integer",
      "default": 16
    },
    "min_citations": {
      "anyOf": [
        {
          "type": "integer"
        },
        {
          "type": "null"
        }
      ],
      "default": null
    },
    "output_format": {
      "type": "string",
      "default": "evidence"
    }
  }
}
search_papers output schema
{
  "type": "object",
  "additionalProperties": true
}
search_papers annotations
{
  "title": "Search scientific evidence",
  "readOnlyHint": true,
  "idempotentHint": true
}
lookup_paper — Look up paper by DOI
Advertised by the server during the last successful anonymous check.
Advertised read-only

Look up a single paper by its DOI. Args: doi: The DOI of the paper (e.g. "10.1038/s41586-024-07386-0"). output_format: "evidence" for compact claim-level evidence (default), "legacy" for original paper metadata, or "full" for both. Returns: An envelope with found status and the paper in result, or a not-found message. A found paper counts as one result.

lookup_paper input schema
{
  "type": "object",
  "required": [
    "doi"
  ],
  "properties": {
    "doi": {
      "type": "string"
    },
    "output_format": {
      "type": "string",
      "default": "evidence"
    }
  }
}
lookup_paper output schema
{
  "type": "object",
  "additionalProperties": true
}
lookup_paper annotations
{
  "title": "Look up paper by DOI",
  "readOnlyHint": true,
  "idempotentHint": true
}

Prompts

0 advertised

No prompt definitions available.

Resources

0 advertised

No resource definitions available.

30-day availability

Aggregated from read-only remote checks; this is not a tool-execution or security guarantee.

Successful checks
1 of 1 (100%)
Average latency
1056 ms
95th percentile latency
1056 ms

Maintenance history

Public check and source-refresh events. Editorial and ownership records remain private.

  • · probe · passed · 1056 ms · metadata changed · 2 tool(s), 0 resource(s), 0 resource template(s), 0 prompt(s)

Connect from your client

Each snippet registers the same endpoint; pick the client you use.

This server is operated by the named third party. Listing and endpoint checks are not an endorsement or a security audit. Review the operator's permissions, terms, and data handling before connecting. How listings are checked and corrected.

Claude Code

Run this once in a terminal. It registers the remote server. Use --scope project to write .mcp.json for the repo.
Claude Code configuration
claude mcp add --transport sse bgpt-mcp https://bgpt.pro/mcp/sse

Cline

Merge this into cline_mcp_settings.json (Cline → MCP Servers → Configure).
Cline configuration
{
  "mcpServers": {
    "bgpt-mcp": {
      "type": "sse",
      "url": "https://bgpt.pro/mcp/sse",
      "disabled": false
    }
  }
}

Gemini CLI

Run this once. It writes .gemini/settings.json in the current project; pass --scope user for ~/.gemini/settings.json.
Gemini CLI configuration
gemini mcp add --transport sse bgpt-mcp https://bgpt.pro/mcp/sse

Let your agent find its tools

Find servers like BGPT MCP from your agent. Add Koskamo once to search the index for endpoints, access requirements, and client setup. Public discovery needs no key.