mcpField
P
Paper Nexus

Paper Nexus

Remote
papernex.usSearch0Updated 2d ago

About Paper Nexus

Paper Nexus is a hosted research discovery service for assistants and researchers. It uses citation links across OpenAlex and Semantic Scholar to recommend related papers, with read-only tools and links to citation graphs.

Key features

  1. 01Search by title, DOI, or link
  2. 02Recommend papers through citation networks
  3. 03Retrieve paper details and cited references
  4. 04List citing papers and library entries
  5. 05Filter results and select returned fields

Use cases

  1. 01Find papers related to selected research
  2. 02Explore references around a key paper
  3. 03Suggest papers missing from a saved library

MCP support

The hosted MCP server at `https://papernex.us/mcp` exposes read-only paper search, recommendation, citation, detail, and library tools. MCP clients connect using OAuth.

Configuration

Remote servers need no local install. Pick your client, then copy the config or install in one click.

Endpointhttps://papernex.us/mcp
Copy install command
Claude CodeTerminal
Cursor.cursor/mcp.json
VS Code.vscode/mcp.json
Claude Desktopclaude_desktop_config.json
Windsurf~/.codeium/windsurf/mcp_config.json

Run commands in a terminal; paste JSON into the client’s MCP config and restart. Servers that need sign-in will prompt on first connect.

Tools

Details

From the project README

Paper Nexus

Paper Nexus for AI assistants

Find related scientific papers through their citation network, from Claude, ChatGPT, Cursor, VS Code or any MCP client.

MCP Badge


Paper Nexus maps the scientific literature by its citations, using the scholarly works in OpenAlex and Semantic Scholar. This repository holds its MCP server's public definition: the connection details, a Claude Code plugin, a Gemini CLI extension, the MCP Registry entry, and the tool reference. The server itself is hosted; there is nothing to install or run.

https://papernex.us/mcp

Ask your assistant things like:

  • "Find papers related to Attention Is All You Need and BERT together."
  • "What are the main references of the AlphaFold 2 paper, and what else should I read around it?"
  • "Starting from https://arxiv.org/abs/1512.03385, which papers on residual networks should I read?"
  • "Go through my Thesis / Methods collection and suggest papers I haven't saved."

Every paper it returns links to papernex.us, and each set of recommendations comes with a link that opens the whole citation graph in Paper Nexus.

Connect

You sign in with a Paper Nexus account the first time you connect; a free account works. There are no API keys: the server uses OAuth, so your assistant opens a Paper Nexus page where you sign in and allow access.

Claude (claude.ai and Claude Desktop)

Settings → Connectors → Add custom connector. Name it Paper Nexus, give the URL https://papernex.us/mcp, then Connect and sign in.

Claude Code
claude mcp add --transport http papernexus https://papernex.us/mcp

Or install the plugin, which adds a skill that teaches Claude a literature-search workflow with these tools:

claude plugin marketplace add PaperNexus/papernexus-mcp
claude plugin install papernexus@papernexus

Then run /mcp, choose papernexus, and sign in.

ChatGPT

Turn on developer mode (Settings → Apps → Advanced settings), then create an app with the URL https://papernex.us/mcp and OAuth authentication, and sign in.

Cursor

Install in Cursor, or add it to ~/.cursor/mcp.json:

{
  "mcpServers": {
    "papernexus": {
      "url": "https://papernex.us/mcp"
    }
  }
}
VS Code

Install in VS Code, or add it to .vscode/mcp.json:

{
  "servers": {
    "papernexus": {
      "type": "http",
      "url": "https://papernex.us/mcp"
    }
  }
}
Gemini CLI
gemini extensions install https://github.com/PaperNexus/papernexus-mcp
Other clients

Any client that speaks MCP over Streamable HTTP with OAuth can connect to https://papernex.us/mcp. The server supports Client ID Metadata Documents and dynamic client registration, and every protocol revision from 2025-03-26 on.

Tools

ToolWhat it does
search_papersFind a paper by its title, a DOI or a link, and get its Paper Nexus id
recommend_papersPapers related to one or more papers through their citation network
get_paperOne paper's details and its most-cited references
get_usageHow many counted calls your account has left
list_libraryYour Paper Nexus library, if you allow "Read your library" when connecting
list_citing_papersThe papers citing one paper

recommend_papers and list_citing_papers also take filters (years, citation counts, authors, venues, title words, citing or cited papers, open access, document types) on every account. Each recommendation says which origin papers it is linked to and how, so your assistant can tell you why a paper is there. Every tool that lists papers takes fields, to return only the details you need, such as titles and years. Arguments and results are in the tool reference. Every tool is read-only.

Allowance

recommend_papers and list_citing_papers count against your allowance. The other tools share a larger daily limit of lookups, and every tool a limit of 30 calls a minute. get_usage tells your assistant how many calls are left.

PlanCounted callsLookupsOrigin papers per recommendation
Free10 a week100 a dayup to 2
Premium200 a day1,000 a dayup to 10

A filtered call is counted like any other. A recommendation or citing list that fails or finds nothing isn't counted. See pricing for Premium.

Privacy and access

  • Connecting asks for your permission on a Paper Nexus page that names the app and where it will send you back. Library access is asked for separately and is read-only.
  • Paper Nexus never sees your conversation, only the tool calls your assistant makes. Recommendation requests are saved to your Paper Nexus history, as on the site.
  • You can see and revoke connected apps in Paper Nexus under Account settings → AI assistants.

Privacy policy · Terms

Support

Open an issue or email [email protected].

License

The files in this repository are MIT licensed; see LICENSE. The Paper Nexus service and its data are covered by its terms.

Comments

Markdown supported

New to MCP?

From what the protocol is to installing your first server, these answers cover most first questions.

An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.