Paper2Agent is a multi-agent AI system designed to automatically convert research papers into interactive AI agents. It targets researchers and developers seeking to leverage complex scientific codebases with minimal human effort, offering a streamlined way to access and utilize the functionalities described in research publications.

How It Works

The system processes a given GitHub repository associated with a research paper, automatically identifying and executing relevant tutorials. It extracts functional tools from these tutorials, packages them into an agent-specific MCP (Model-centric Pipeline) server, and integrates with AI coding assistants like Claude Code. This approach aims to create interactive, executable agents directly from research artifacts.

Quick Start & Requirements

• Installation: Clone the repository (git clone https://github.com/jmiao24/Paper2Agent.git), navigate into it (cd Paper2Agent), install dependencies (pip install fastmcp), and install Claude Code (npm install -g @anthropic-ai/claude-code).
• Prerequisites: Python 3.10+ and Claude Code.
• Runtime: Processing can take 30 minutes to over 3 hours depending on codebase complexity.
• Cost: Estimated at ~$15 for complex repositories using specific models (e.g., Claude Sonnet 4).
• Usage: Initiated via bash Paper2Agent.sh with --project_dir and --github_url. Tutorials can be filtered using --tutorials, and API keys can be provided via --api.
• Documentation: Demos and connectable MCP servers are linked within the README.

Highlighted Details

• Automates the creation of interactive AI agents from research papers.
• Extracts and operationalizes code tutorials into reusable tools.
• Supports specialized agents for genomics (AlphaGenome), spatial transcriptomics (TISSUE), and single-cell analysis (Scanpy).
• Generates isolated Python environments and detailed project outputs, including MCP servers and test suites.

Maintenance & Community

The provided README does not detail specific community channels (e.g., Discord, Slack), active maintainers, or corporate sponsorships. It includes an academic citation for the work.

Licensing & Compatibility

The repository's license is not specified in the README. This lack of clarity may pose a barrier for commercial use or integration into closed-source projects.

Limitations & Caveats

Processing times can be lengthy, ranging from 30 minutes to several hours. The system incurs costs, estimated at $15 for complex projects. Reliance on external AI coding assistants like Claude Code is a key dependency. The absence of a specified license is a significant caveat for adoption.