Paper2Agent: Stanford's Framework for Turning Research Papers into Interactive AI Agents

Forum topic · ✨步子哥 · 2025-10-07

Summary

Paper2Agent is an automated framework proposed by Stanford University researchers that converts scientific papers into 'research assistant' AI agents. Built on the Model Context Protocol (MCP) ecosystem, it has two key components: Paper2MCP, which packages a paper's methodological contributions into MCP servers, and an agent layer that connects them to LLM agents. Each MCP server exposes tools (executable functions encapsulating the paper's methods), resources (paper text, code repositories, and supplementary materials), and prompts (instruction templates for multi-step scientific workflows). The conversion pipeline identifies a paper's key contributions (datasets, methods, models, or workflows), wraps them into an MCP server with defined inputs, outputs, and usage instructions, and connects it to an LLM agent for natural-language queries and autonomous execution. The resulting paper agents support interactive, reproducible scientific analysis without programming, validated against example datasets with locked configurations, and enable multi-agent collaboration. A showcased case study is the AlphaGenome agent. The framework aims to shift scientific research from 'reading' papers to 'conversing' with them, allowing published work to be directly invoked as usable tools.

Paper2Agent: Turning Papers into Interactive AI Agents (Stanford)

Overview

Paper2Agent is an automated framework proposed by Stanford University researchers that transforms scientific papers into "research assistant" style AI agents. Published academic results become directly invocable, moving research from *reading* papers to *conversing* with them. The framework supports multi-agent collaboration.

Characteristics of paper agents

Easy interaction: Perform complex scientific analyses via natural language, no programming required.
Reliable and reproducible: Tools are validated against example datasets, with locked configurations to ensure reproducibility.

Technical Implementation

Paper2Agent is built on the MCP (Model Context Protocol) ecosystem, with two key components: Paper2MCP and the agent layer.

Core components of an MCP server

MCP Tools: Executable functions that encapsulate the methodological innovations of the paper.
MCP Resources: Static repositories containing the paper text, code repositories, and supplementary materials.
MCP Prompts: Refined instruction templates that guide AI agents through multi-step scientific workflows.

Conversion pipeline

1. Identify key contributions — datasets, methods, models, or workflows. 2. Wrap as an MCP server — with explicit inputs, outputs, and usage instructions. 3. Connect to an LLM agent — enabling natural-language queries and autonomous execution.

Case Study: AlphaGenome Agent

The post showcases an AlphaGenome agent built with Paper2Agent as its primary case study, demonstrating how a published genomic model can be exposed as an interactive tool for natural-language scientific analysis.

Why It Matters

By packaging papers as standards-compliant MCP servers, Paper2Agent makes published methods immediately usable, reproducible, and composable — turning the scientific literature into a live ecosystem of callable AI tools.