FreeHow credits work →

Research tools for Kimi Code

Add ScholarMuse with /mcp-config and log in from the same prompt. Kimi Code can then find the paper you want to reproduce and fetch the catalogues and observations its analysis needs.

Connect your agentSee all capabilities
ClaudeCodexChatGPTCursorGemini CLIAntigravityWorkBuddyTraeCodeBuddyCherry StudioQoderQwen CodeKimi Code

How to connect ScholarMuse to Kimi Code

The steps follow the Kimi Code documentation and cover the terminal agent and the VS Code extension. Kimi Code runs shell commands and fetches web pages, so pasting “Read https://scholarmuse.ai/install/agent.md and follow the instructions to connect ScholarMuse” should also work.

Full connect guide →

  1. Step 1

    Terminal: add the server

    Type /mcp-config in Kimi Code and add the server there, or put this in ~/.kimi-code/mcp.json and keep any entries that are already there. An entry with a url and no transport is treated as Streamable HTTP, which is what ScholarMuse uses.

    {
  "mcpServers": {
    "scholarmuse": {
      "url": "https://scholarmuse.ai/mcp"
    }
  }
}

  2. Step 2

    Terminal: log in

    Run this inside Kimi Code. It starts the browser authorization for the server you just added.

    /mcp-config login scholarmuse

  3. Step 3

    VS Code extension: add an http server

    Click the gear icon, open MCP Servers, choose the type http and enter this address. Then use the authorize button next to the server.

    https://scholarmuse.ai/mcp

  4. Step 4

    Approve in your browser

    A scholarmuse.ai window opens. Sign in, approve access and choose which discipline packs Kimi Code may use. There is no API key to create or paste.

  5. Step 5

    Check that it works

    Ask Kimi Code to list the ScholarMuse capabilities and show three things it can now do for your research. If the tools do not appear, restart Kimi Code and ask again.

What Kimi Code can use to reproduce a result

Ordered the way a reproduction goes: find the paper and its data first, then the catalogues and observation series from astronomy and earth science, then a last check before the notebook is shared.

Search papers

Search papers

All disciplines · OpenAlex, PubMed / Europe PMC, arXiv +5

Find datasets

Find datasets

All disciplines · DataCite, Zenodo, Dryad +2

Exoplanets & stars

Exoplanets & stars

Astronomy · NASA Exoplanet Archive, Gaia Archive, SIMBAD

Earthquake & water data

Earthquake & water data

Earth & environment · USGS

Historical weather data for any placeNew

Historical weather data for any place

Earth & environment · NASA POWER

Compounds & materials

Compounds & materials

Chemistry & materials · PubChem, ChEBI, Materials Project +2

Economic & social statistics

Economic & social statistics

All disciplines · FRED, World Bank, IMF +5

Retraction checker

Retraction checker

All disciplines · Retraction Watch, Crossref

Search papers
Start from the paper itself: its record, DOI and an open-access copy where one exists, so Kimi Code works from the published method.
Find datasets
Check whether the authors deposited their data, and under which licence, before rebuilding it from raw sources.
Exoplanets & stars
Query the NASA Exoplanet Archive and the Gaia catalogue for the sample a figure was drawn from.
Earthquake & water data
USGS earthquakes around a point for a chosen period (up to 200 events per call), or discharge at a stream gauge, as the raw table of a geoscience plot.
Historical weather data for any place
Daily temperature, precipitation and solar radiation at given coordinates, for trend estimates and forcing data.
Compounds & materials
Reference values for compounds and crystals to compare with the numbers your own calculation produces.
Economic & social statistics
Official economic and social series when the result to reproduce is an empirical table.
Retraction checker
Run the DOIs cited in a notebook or README through a retraction check before you publish the repository.

Things to ask Kimi Code

Search papers

You say to your agent

Use ScholarMuse to find the 2017 paper by Fulton and colleagues that reported a gap in the radius distribution of small planets. Give me its DOI and an open-access link, and summarise the sample cuts in notes/method.md.

The matching record with DOI, authors, venue and an open-access link where one exists.

Exoplanets & stars

You say to your agent

Query the NASA Exoplanet Archive for planets within 100 parsecs with radii below 4 Earth radii. Save the table as data/planets.csv and plot the radius histogram on a log axis.

A table of planets from the NASA Exoplanet Archive that Kimi Code writes to disk and plots.

Earthquake & water data

You say to your agent

Get the earthquakes of magnitude 6 and above within 300 km of 38.3 N, 142.4 E during March 2011, and plot magnitude against time to show the aftershock sequence.

A USGS catalogue with time, location, depth and magnitude for each event, ready for the plotting code.

Frequently asked questions

No key. ScholarMuse has its own sign-in at scholarmuse.ai, which opens in your browser when you run /mcp-config login. You approve access there and nothing is pasted into Kimi Code. New ScholarMuse accounts get 50 credits once.

The address and the account are the same; the place differs. In the terminal you use /mcp-config or ~/.kimi-code/mcp.json and then /mcp-config login scholarmuse. In the VS Code extension you add an http server under MCP Servers and use its authorize button.

ScholarMuse supplies the inputs: the paper record, deposited datasets, catalogues and observation series, each with its source and licence. Writing and running the analysis is Kimi Code’s work, and judging whether the reproduced numbers match the paper stays with you.

Every call your agent makes uses credits, and a call is charged only when it succeeds. There are no subscriptions. A new account starts with a one-time gift of 50 credits. The 6 web tools under /tools are free without an account, up to 5 runs a day.

Requests are used only to answer you. We do not train models on them.

Ready to make your agent a better scholar?

Connect once. Papers, data and figures from 50+ sources, in the agent you already use.

Works with compatible AI agents

ClaudeChatGPTCodexCursorWorkBuddyTraeGemini CLIAntigravityQoderCodeBuddyKimi CodeQwen CodeClineKiroWindsurfZedOpenCodeOpenClawDifyCherry StudioLM StudioGooseAmpContinueRoo CodeVS CodeJetBrainsWarpRaycastCopilotPerplexityAny MCP client

Data from OpenAlex · Crossref · PubMed · arXiv · NIH · UniProt · PDB · PubChem · NASA · World Bank · USGS · GBIF and more