Research Methods Library

A Model Context Protocol server that grounds AI assistants in 12,921 sections of methodological texts on causal inference, econometrics, experimental design, marketing research, credit scoring, and artificial intelligence. Responses include study design recommendations, the assumptions to verify, runnable R and Python code, and a citation for every source. Access is free and requires no account or API key.

https://mcp.research-methods.deno.net/mcp

Add to Cursor Install in VS Code

Installation

Claude (web, desktop, and mobile)

Settings → Connectors → Add custom connector; enter the endpoint URL above.

Claude Code

claude mcp add --transport http research-methods https://mcp.research-methods.deno.net/mcp

ChatGPT

Enable developer mode in Settings, add a custom connector of type MCP server with the endpoint URL above, and select no authentication.

Cursor, Visual Studio Code, Windsurf, Cline, Zed, and other MCP clients

Use the buttons above, or add the following to the client's MCP configuration:

{
  "mcpServers": {
    "research-methods": {
      "url": "https://mcp.research-methods.deno.net/mcp"
    }
  }
}

Local installation (requires Node.js 20 or later)

{
  "mcpServers": {
    "research-methods": {
      "command": "npx",
      "args": [
        "-y",
        "research-methods-mcp"
      ]
    }
  }
}

Example queries

Study design

"States raised the minimum wage in different years. How do I estimate the effect on teen employment?"

Implementation

"Give me Python code for a regression discontinuity with a data-driven bandwidth."

Identifying assumptions

"What could break my synthetic control estimate, and how do I test for it?"

Citation

"Write my methods section and give me BibTeX for the sources."

Tools

choose_method: Matches a research question to methods, with assumptions, R and Python packages, and readings.
search and fetch: Find and read sections, with code and links. Compatible with ChatGPT deep research.
get_code: Runnable R, Python, or Julia code for a task.
cite: References in APA, Chicago, MLA, Harvard, BibTeX, or RIS for the sections consulted.
list_sources: The books and packages in the library.

Sources

A Guide on Data Analysis

Data analysis, econometrics, and causal inference in R: regression, GLMs, mixed models, experiments, difference-in-differences, synthetic control, regression discontinuity, instrumental variables, matching, double machine learning, mediation, sensitivity analysis, and structural IO.

Bash, Python, R · 1,611 sections

causalverse: The All-in-One Causal Inference Toolkit

R package for the full causal inference pipeline: RCTs and power, DiD and staggered adoption, synthetic DiD, synthetic control, RD, IV, event studies, matching, heterogeneous effects, mediation, sensitivity, and publication-ready output. Vignettes plus a function reference.

R · 1,163 sections