Lit Review Agent Tools

Awesome AI Literature Review [](https://awesome.re)

> Open-source AI agents, skills, MCP servers, and frameworks for literature review.

Curated tools that help AI agents handle the whole loop of a literature review: search, screen, extract, read, synthesize, verify citations, and write.

Every entry carries a maintenance status and a licence, refreshed weekly from the GitHub API, because a dead link and a non-commercial licence both cost you more than a missing star count. Four of the workflows here were run end to end rather than described.

• 🔍 Searchable version — filter by workflow stage, maintenance status and licence
• 🧪 [Verified workflows](recipes/) — executed, with real output and what broke
• 🧭 [Browse by workflow stage](STAGES.md) — search · screen · extract · read · synthesize · cite-check · write · review
• 🎯 [The ten picks, and when not to use them](PICKS.md)
• 📊 [Health report](HEALTH.md) · [dataset](data/tools.json) · [changelog](CHANGELOG.md)
• 🇨🇳 [简体中文版](README.zh.md) — 完整中文表格版

Contents

• [All-in-one Research Agents and Skills](#all-in-one-research-agents-and-skills)
• [Deep Research and Auto Survey Generation](#deep-research-and-auto-survey-generation)
• [Autonomous Science](#autonomous-science)
• [Paper Q and A and RAG](#paper-q-and-a-and-rag)
• [Systematic Review and Screening](#systematic-review-and-screening)
• [MCP Servers](#mcp-servers)
• [Reference and Knowledge Management](#reference-and-knowledge-management)
• [PDF to Structured Data](#pdf-to-structured-data)
• [Citation Graphs and API Clients](#citation-graphs-and-api-clients)
• [Writing and Peer Review](#writing-and-peer-review)
• [Awesome Lists](#awesome-lists)

All-in-one Research Agents and Skills

• academic-research-skills - Claude Code skills running a ten-stage research-to-paper pipeline with citation integrity gates.
• academic-research-skills-codex - Codex-native sibling with a human-in-the-loop research flow.
• Research-Paper-Writing-Skills - Machine-learning paper-writing skill pack for Codex, Claude Code, and Gemini.
• claude-skills - Large skill collection including a literature-review, grants, and deep-research stack.
• academic-paper-skills - Planning and writing skills with quality checkpoints for systematic paper writing.
• dr-claw - Research workbench with multiple AI-assistant personas.
• ScienceClaw - Self-evolving AI research colleague with a large skill library.
• qinyan-academic-skills - Multilingual library of installable AI-agent skills across disciplines.
• agent-research-skills - Claude Code skills for systematic literature review with citation-validation scripts.
• medsci-skills - Medical-research skills for search, reporting-guideline checks, statistics, and submission.

Deep Research and Auto Survey Generation

• STORM - Retrieval-grounded system that writes cited, encyclopedia-style long articles from a topic.
• gpt-researcher - Autonomous agent that runs deep research on any topic and outputs a cited report.
• deep-research - Minimal iterative deep-research agent that is small and easy to hack on.
• opendeep_research - Open deep-research reference implementation from LangChain.
• local-deep-research - Local and private deep-research over many sources including arXiv and PubMed.
• open-deep-research - Open deep-research clone that reasons over web data via Firecrawl.
• SurveyX - Automated generation of academic survey papers from a topic.
• LitLLM - Toolkit that drafts related-work sections fast with retrieval and prompting.
• opendraft - Free and open-source AI paper writer where many agents collaborate on long drafts.
• AutoSurveyGPT - Finds and ranks scholarly papers and auto-generates a survey.

Autonomous Science

• AI-Scientist - End-to-end automated discovery covering literature review, experiments, writing, and review.
• AI-Scientist-v2 - Workshop-level automated scientific discovery via agentic tree search.
• AutoResearchClaw - Self-evolving research from idea to conference-ready paper with real literature and sandboxed experiments.
• Agent-Laboratory - Autonomous workflow from literature review through experimentation to report writing.
• SciAgentsDiscovery - Multi-agent system that generates hypotheses and scientific discoveries.
• Zochi - Artificial-scientist system doing end-to-end discovery toward publication.
• DeepInnovator - Autonomously generates research ideas, hypotheses, and experiment designs.

Paper Q and A and RAG

• paper-qa - High-accuracy retrieval system for scientific papers whose answers always cite sources.
• paperai - Semantic search and question answering over medical and scientific papers.
• openpaper - Research-library workbench with a grounded, citation-aware review assistant.

Systematic Review and Screening

• ASReview - Active-learning screener that ranks papers to cut systematic-review screening time.
• LatteReview - Low-code package that automates systematic-review screening with AI agents.
• prismAId - Protocol-based, no-code toolkit for replicable review screening and extraction.
• prisma-review-tool - Structured review flow with AI-assisted screening over open scholarly sources.

MCP Servers

• zotero-mcp - Connects a Zotero library to AI clients for semantic search and citation analysis.
• arxiv-mcp-server - Searches and analyzes arXiv papers and converts them to Markdown for context.
• paper-search-mcp - Multi-source search and download across many scholarly databases.
• PubMed-MCP-Server - Searches, accesses, and analyzes PubMed articles through a tool interface.
• alex-mcp - OpenAlex server focused on author disambiguation and institution lookup.
• openalex-research-mcp - OpenAlex-backed citation analysis, trend tracking, and collaboration mapping.

Reference and Knowledge Management

• zotero-gpt - Integrates a chat assistant into Zotero to talk with your library.
• papersgpt-for-zotero - Zotero plugin to chat with and batch-process PDFs across many models.
• ai-research-assistant - Assistant that lives inside Zotero for research tasks.
• paper-note-filler - Obsidian plugin that auto-creates notes from scholarly sources.

PDF to Structured Data

• MinerU - High-accuracy conversion of PDFs and office files to LLM-ready Markdown and JSON.
• docling - Document parser that prepares PDFs and documents for generative AI and RAG.
• marker - Fast conversion of PDFs and documents to clean Markdown and JSON.
• PDFMathTranslate - Layout-preserving scientific-PDF translation that keeps formulas and figures intact.
• grobid - Extracts structured metadata, references, and sections from scholarly PDFs.
• paperetl - ETL pipeline that loads medical and scientific papers into structured stores.
• scipdf_parser - Python parser for scientific PDFs that extracts content and figures.

Citation Graphs and API Clients

• scholarly - Pythonic retrieval of authors and publications from Google Scholar.
• semanticscholar - Python client for the Semantic Scholar APIs.
• pyalex - Lightweight Python interface to the OpenAlex API.
• ArxivDigest - Personalized daily arXiv digest with relevancy scoring and email delivery.
• citegraph - Open web visualizer of papers and citation networks.

Writing and Peer Review

• lmms-lab-writer - Local-first agentic LaTeX writer for AI-assisted academic writing.
• academic-writing-agents - Claude Code plugin with specialist agents for reviewing, drafting, and polishing.
• ai-peer-review - Multi-model meta-review that synthesizes independent reviews of a paper.
• openreviewer - Generates peer reviews of conference papers for pre-submission feedback.
• academic-research-plugin - Claude Code plugin for surveys, paper reviews, and finding research gaps.

Awesome Lists

• Awesome-LLM-Scientific-Discovery - Survey list of language models in scientific discovery.
• Awesome-Auto-Research-Tools - Automated tools for search, reading, experiments, and code generation.
• awesome-ai-auto-research - Survey of AI approaches to automated research.
• LLM4SR - Papers and resources on language models for scientific research surveys.
• awesome-ai-research-tools - AI tools for literature reviews, reference management, and data analysis.
• awesome-evidence-synthesis - Open-source tools for systematic reviews and meta-analysis.

Contributing

Contributions welcome. Read the [contributing guidelines](contributing.md) first.

Source & license

This open-source MCP server is cataloged on AgentStack and links to its original source — we do not rehost the code.

• Author: brycewang-stanford
• Source: brycewang-stanford/lit-review-agent-tools
• License: CC0-1.0
• Homepage: https://brycewang-stanford.github.io/lit-review-agent-tools/

Install and usage instructions live in the source repository linked above.