Paper2Agent is a multi-agent AI system that automatically transforms research papers into interactive AI agents.
The tool solves the problem of making research implementations accessible and executable without manual engineering effort. It works by coordinating parallel specialist agents that analyze a research paper alongside its source code repository, extracting key concepts and functionality to generate reliable MCP servers and reusable skills. Users provide minimal input—typically just a paper URL and optional code repository—and the system handles the transformation automatically.
Paper2Agent suits researchers and developers who want to quickly operationalize published methods or create interactive tools from existing codebases. It works best when the paper has an associated public repository with working code. The tool is designed as a skill that integrates with coding agents like Claude Code or Codex, allowing users to invoke it directly from their development environment. It supports specifying targeted tasks or tutorial URLs to focus the agent generation on particular aspects of a paper, and it handles credential management by keeping API keys and secrets outside generated code and deliverables.
The project shows active development with ongoing refinement of the multi-agent coordination system. The codebase demonstrates attention to practical deployment concerns, including isolated project environments for generated agents and version tracking for dependencies. The tool maintains clear separation between the coordinator agent and specialist agents, with mechanisms for spawning fresh verifier agents to validate outputs.
