OpenReview

OpenReview

An MCP server for OpenReview to fetch, read and save manuscripts from AI/ML conferences.

2
Stars
1
Forks
1
Releases

Overview

The OpenReview MCP server is an MCP service that provides programmatic access to OpenReview data for research and analysis. It supports user search by email address, enabling discovery of OpenReview profiles; paper retrieval by author to fetch all publications; and conference papers queries to retrieve papers from major ML venues (ICLR, NeurIPS, ICML) by venue and year. It also offers keyword-based search across venues and supports exporting results as JSON and PDFs for offline reading, data analysis, and aiding coding assistants. The server is configured via a .env file with OpenReview credentials and optional settings for the OpenReview base URL and default export directory. It includes example usage with Claude Code, showing how to add the MCP server to a Claude Code project and run queries; it exposes a set of tools for common tasks: search_user, get_user_papers, get_conference_papers, search_papers, and export_papers. Supported conferences are ICLR, NeurIPS, and ICML.

Details

Owner
anyakors
Language
Python
License
MIT License
Updated
2025-12-07

Features

User search

Find OpenReview profiles by email address.

Paper retrieval

Fetch all papers by a specific author.

Conference papers

Get papers from specific venues (ICLR, NeurIPS, ICML) and years.

Keyword search

Search papers by keywords across multiple conferences.

JSON & PDF export

Export search results to PDF and JSON files for convenient reading or further analysis and coding assistant usage.

Audience

ResearchersSearch OpenReview profiles, fetch author papers, and export results for research and analysis.
Data scientistsAggregate conference papers across venues to support ML research tasks and analyses.

Tags

OpenReviewMCPresearchpapersconference_papersICLRICMLNeurIPSuser_searchexportJSONPDF