Content
<div align="center">
<img src="assets/banner.svg" alt="Awesome AI Literature Review Agents & Tools" width="100%">
# Tool List
**🤖 Awesome AI Literature Review Agents & Tools**
_Allow AI agents to help you complete the entire process of literature review: retrieval → reading → extraction → review → citation verification._
[](https://awesome.re)
[](CONTRIBUTING.md)
[](LICENSE)
[](https://github.com/brycewang-stanford/lit-review-agent-tools/stargazers)
[](https://github.com/brycewang-stanford/lit-review-agent-tools/commits)
**English** · [简体中文](README.md)
<em>Curates <b>70+</b> open-source projects for literature review, categorized by use cases · Updated quarterly · PRs welcome</em>
</div>
---
> Curates open-source AI literature review agents, Claude Code / Codex skills, MCP servers, systematic review screening tools, PDF parsing, citation graph, and automated review frameworks.
> Goal: Become a one-stop navigation for researchers to use AI for literature work. ⭐ **Star and bookmark for future reference.**
## 📑 Table of Contents
- [Why this list](#why-this-list)
- [⚡ 30-second selection](#-30-second-selection)
- [🌟 Comprehensive research agents and skills](#-comprehensive-research-agents-and-skills)
- [🔎 In-depth research and automated review generation](#-in-depth-research-and-automated-review-generation)
- [🧪 Autonomous research: from idea to paper](#-autonomous-research-from-idea-to-paper)
- [📚 Literature question answering and RAG](#-literature-question-answering-and-rag)
- [🧮 Systematic review and literature screening](#-systematic-review-and-literature-screening)
- [🔌 MCP servers](#-mcp-servers)
- [🗂️ Literature management and Zotero / Obsidian integration](#️-literature-management-and-zotero--obsidian-integration)
- [📄 PDF → structured data extraction](#-pdf--structured-data-extraction)
- [🕸️ Citation graph and API clients](#️-citation-graph-and-api-clients)
- [✍️ Paper writing and peer review assistants](#️-paper-writing-and-peer-review-assistants)
- [📖 Awesome lists](#-awesome-lists)
- [🏢 Commercial / closed-source tools (reference)](#-commercial--closed-source-tools-reference)
- [🧭 How to choose (decision table)](#-how-to-choose-decision-table)
- [📈 Star trends](#-star-trends)
- [🤝 Contributing](#-contributing)
- [📄 License](#-license)
> ⭐ = Editor's recommendation ｜ Stars are approximate values from GitHub API, **automatically updated weekly**, for reference only.
---
## Why this list
Literature review is one of the most time-consuming parts of research: retrieval, deduplication, in-depth reading, theme induction, citation verification...
Recently, many AI tools have emerged that can automate some (or all) of these steps, but they are scattered across GitHub, vary in quality, and have confusing names.
This list follows three principles:
1. **Categorized by use cases** — Come with a specific goal in mind, not searching through hundreds of repositories.
2. **Only strong and usable open-source projects** — Closed-source tools are listed separately for reference.
3. **Accurate information** — Star numbers are verified with GitHub API, and links are continuously checked with CI.
---
## ⚡ 30-second selection
```
I use Claude Code and want an end-to-end "research → paper" solution ─────────▶ academic-research-skills ⭐
I want AI to automatically survey a topic and generate a report with citations ─────────▶ GPT Researcher / STORM
I want a fully automated "from idea to publishable paper" solution ───────────────────▶ AI-Scientist-v2 / AutoResearchClaw
I want to ask questions about a batch of PDFs with citations ───────────────────▶ PaperQA2
I'm doing a rigorous PRISMA systematic review (screening thousands of abstracts) ──────────▶ ASReview / prismAId
I want to convert PDFs into clean Markdown for LLMs ────────────▶ MinerU / Docling / marker
I want to connect literature capabilities to Claude / Cursor (MCP) ──────────▶ paper-search-mcp / zotero-mcp
I want to converse with my literature library directly in Zotero ───────────────────▶ zotero-gpt / PapersGPT
```
---
## 🌟 Comprehensive research agents and skills
> End-to-end solutions covering the entire research → writing → review → revision process, mostly Claude Code / Codex skills.
| Project | Stars | Description |
|---|---|---|
| ⭐ [academic-research-skills](https://github.com/Imbad0202/academic-research-skills) | ~39.5k | **Most popular project in this field.** A set of Claude Code skills with a 10-stage pipeline (research → writing → review → revision → finalization), including a built-in citation/integrity "gate", cross-referencing Semantic Scholar + OpenAlex + Crossref. The idea is * "AI is a co-pilot, not the pilot"* . `/plugin install academic-research-skills` |
| [academic-research-skills-codex](https://github.com/Imbad0202/academic-research-skills-codex) | ~7.1k | Codex-native version of the above, with human-in-the-loop academic research workflow |
| [Research-Paper-Writing-Skills](https://github.com/Master-cai/Research-Paper-Writing-Skills) | ~5.5k | ML/CV/NLP paper writing skills package (adapted from Peng Shida's notes), compatible with Codex / Claude Code / Gemini |
| [claude-skills](https://github.com/alirezarezvani/claude-skills) | ~23.2k | Large skill collection, including litreview / grants / deep-research stacks, across Claude Code / Codex / Gemini / Cursor |
| [academic-paper-skills](https://github.com/lishix520/academic-paper-skills) | ~1.1k | Strategist (planning) + Composer (writing) dual skills with quality checkpoints |
| [dr-claw](https://github.com/OpenLAIR/dr-claw) | ~1.0k | "Research IDE" with multiple AI assistant roles |
| [ScienceClaw](https://github.com/beita6969/ScienceClaw) | ~0.9k | Self-evolving AI research companion with 285 skills, focusing on "zero hallucination" |
| [qinyan-academic-skills](https://github.com/LeonChaoX/qinyan-academic-skills) | ~0.7k | Multi-language library of 182 installable AI agent skills covering multiple disciplines |
| [agent-research-skills](https://github.com/lingzhi227/agent-research-skills) | ~0.2k | Claude Code skills for systematic literature review with citation verification scripts |
| [medsci-skills](https://github.com/Aperivue/medsci-skills) | ~0.2k | Medical research skills: retrieval, report specification/citation verification, statistics, publication charts, and submission (produced by doctors-researchers) |
---
## 🔎 In-depth research and automated review generation
> Input a topic, automatically retrieve and generate a review/report/related work section with citations.
| Project | Stars | Description |
|---|---|---|
| ⭐ [STORM](https://github.com/stanford-oval/storm) | ~30.3k | Stanford OVAL's product, based on retrieval, a two-stage process ("pre-writing" + "writing") generating Wikipedia-style long articles with citations; includes a multi-agent conversational version Co-STORM |
| ⭐ [gpt-researcher](https://github.com/assafelovic/gpt-researcher) | ~28.6k | Autonomous research agent, conducting in-depth research on any topic and generating reports with citations; general-purpose, not limited to academia |
| [deep-research](https://github.com/dzhng/deep-research) | ~19.4k | Minimalist iterative in-depth research agent (search + crawl + LLM refinement), short and easy to modify |
| [open_deep_research](https://github.com/langchain-ai/open_deep_research) | ~12.4k | LangChain's official open-source reference implementation for deep research |
| [local-deep-research](https://github.com/LearningCircuit/local-deep-research) | ~8.8k | Local/privacy-first deep research with 10+ data sources including arXiv / PubMed, fully local LLM support |
| [open-deep-research](https://github.com/nickscamara/open-deep-research) | ~6.3k | Open-source deep research recreation based on Firecrawl's web data inference |
| [SurveyX](https://github.com/IAAR-Shanghai/SurveyX) | ~1.0k | Automatically generates academic review papers from a topic |
| [LitLLM](https://github.com/LitLLM/LitLLM) | ~44 | Scientific literature review toolkit using RAG + prompt engineering to quickly generate related work (TMLR 2025) |
| [opendraft](https://github.com/federicodeponte/opendraft) | ~0.3k | Free and open-source AI paper writing with 19 agents collaborating to draft long papers |
| [AutoSurveyGPT](https://github.com/a554b554/AutoSurveyGPT) | ~0.2k | Uses GPT to retrieve and rank papers from Google Scholar, automatically generating literature reviews |
---
## 🧪 Autonomous research: from idea to paper
> End-to-end "fully automated scientific discovery" — literature survey + hypothesis generation + experiment running + writing + self-review. Ambitious.
| Project | Stars | Description |
|---|---|---|
| ⭐ [AI-Scientist](https://github.com/SakanaAI/AI-Scientist) | ~14.3k | Sakana AI, end-to-end fully automated scientific discovery (literature → experiment → writing → review); also [v2](https://github.com/SakanaAI/AI-Scientist-v2) ( ~6.9k, agentic tree search, reaching workshop level) |
| [AutoResearchClaw](https://github.com/aiming-lab/AutoResearchClaw) | ~13.9k | Self-evolving autonomous research: from idea to publishable LaTeX paper (OpenAlex/S2/arXiv real literature + sandbox experiment + multi-agent review) |
| [Agent-Laboratory](https://github.com/SamuelSchmidgall/AgentLaboratory) | ~5.8k | End-to-end autonomous research workflow: literature review → experiment → report writing |
| [SciAgentsDiscovery](https://github.com/lamm-mit/SciAgentsDiscovery) | ~0.6k | Multi-agent (ontologist/scientist/critic) automatically generating hypotheses and scientific discoveries |
| [Zochi](https://github.com/IntologyAI/Zochi) | ~0.3k | "Artificial scientist", from discovery to peer-reviewed acceptance |
| [DeepInnovator](https://github.com/HKUDS/DeepInnovator) | ~0.3k | Autonomously generates research ideas, questions, testable hypotheses, and experimental designs |
---
## 📚 Literature question answering and RAG
> For a batch of PDFs/papers, perform **citation-based** question answering and information extraction.
| Project | Stars | Description |
|---|---|---|
| ⭐ [paper-qa](https://github.com/Future-House/paper-qa) | ~8.9k | FutureHouse's product, high-precision scientific literature RAG, answers **always include citations**; PaperQA2 claims to achieve superhuman literature retrieval |
| [paperai](https://github.com/neuml/paperai) | ~1.8k | Semantic search + question answering for medical and research papers |
| [openpaper](https://github.com/khoj-ai/openpaper) | ~0.4k | Research library workspace: reading/annotating papers + AI review assistant with citation tracking |
---
## 🧮 Systematic review and literature screening
> For rigorous evidence-based/PRISMA systematic reviews, efficiently screening from thousands of abstracts.
| Project | Stars | Description |
|---|---|---|
| ⭐ [ASReview](https://github.com/asreview/asreview) | ~1.0k | Active learning-based systematic review screening tool, interactively sorting papers, significantly reducing screening time; mature academic solution |
| [LatteReview](https://github.com/PouriaRouzrokh/LatteReview) | ~0.1k | Low-code Python package using AI agents for automated systematic review screening (OpenAI/Gemini/Claude/Ollama) |
| [prismAId](https://github.com/Open-and-Sustainable/prismAId) | ~24 | Generative AI-based protocol-driven systematic review tool, no programming required, reproducible screening and extraction |
| [prisma-review-tool](https://github.com/Black-Lights/prisma-review-tool) | niche | PRISMA 2020 full process, using MCP for AI-assisted screening (arXiv/OpenAlex/S2, no API key required) |
---
## 🔌 MCP servers
> Connect literature capabilities to Claude / Cursor / Cline, etc., supporting Model Context Protocol clients.
| Project | Stars | Description |
|---|---|---|
| ⭐ [zotero-mcp](https://github.com/54yyyu/zotero-mcp) | ~4.4k | Connects Zotero literature library (local + Web API) to AI: semantic search, PDF full text, citation analysis. Most popular Zotero MCP |
| ⭐ [arxiv-mcp-server](https://github.com/blazickjp/arxiv-mcp-server) | ~3.0k | Retrieves and analyzes arXiv papers, downloads and converts PDFs to Markdown for LLMs; provides `.mcpb` package |
| [paper-search-mcp](https://github.com/openags/paper-search-mcp) | ~2.3k | Searches/downloads papers across 20+ sources (arXiv, PubMed, bioRxiv, S2, OpenAlex, Crossref, CORE…) |
| [PubMed-MCP-Server](https://github.com/JackKuo666/PubMed-MCP-Server) | ~0.1k | Retrieves, accesses, and analyzes PubMed articles (metadata + in-depth analysis) |
| [alex-mcp](https://github.com/drAbreu/alex-mcp) | ~50 | OpenAlex MCP, focusing on author disambiguation and institution/achievement query |
| [openalex-research-mcp](https://github.com/oksure/openalex-research-mcp) | ~37 | OpenAlex (2.4M+ outcomes): citation analysis, research trends, collaboration networks |
---
## 🗂️ Literature management and Zotero / Obsidian integration
> Embed AI into your existing literature management/note-taking workflow.
| Project | Stars | Description |
|---|---|---|
| ⭐ [zotero-gpt](https://github.com/MuiseDestiny/zotero-gpt) | ~7.3k | Integrates GPT into Zotero, directly conversing with your literature library |
| [papersgpt-for-zotero](https://github.com/papersgpt/papersgpt-for-zotero) | ~2.6k | Zotero AI + MCP plugin, chatting/processing PDFs across 30+ LLMs |
| [ai-research-assistant](https://github.com/lifan0127/ai-research-assistant) | ~1.7k | "Aria" LLM research assistant within Zotero |
| [paper-note-filler](https://github.com/chauff/paper-note-filler) | ~47 | Obsidian plugin, automatically building notes from arXiv / ACL Anthology / Semantic Scholar |
---
## 📄 PDF → Structured Data Extraction
> The invisible infrastructure of literature review: turning PDFs into clean, structured Markdown / JSON for LLM.
| Project | Stars | Description |
|---|---|---|
| ⭐ [MinerU](https://github.com/opendatalab/MinerU) | ~75.7k | High-precision PDF/Office → LLM-ready Markdown/JSON (VLM+OCR, 100+ languages, formulas/tables) |
| [docling](https://github.com/docling-project/docling) | ~63.8k | IBM's document parser for GenAI/RAG-ready PDFs/documents |
| [marker](https://github.com/datalab-to/marker) | ~37.8k | Quickly convert PDFs/documents into clean Markdown/JSON, friendly for research documents |
| [PDFMathTranslate](https://github.com/Byaidu/PDFMathTranslate) | ~35.8k | Translate research PDFs with preserved layout (formulas/charts unchanged) |
| [grobid](https://github.com/grobidOrg/grobid) | ~5.0k | Extract structured TEI/XML from academic PDFs (metadata, references, sections) |
| [paperetl](https://github.com/neuml/paperetl) | ~0.7k | ETL pipeline for medical and research papers, storing structured data |
| [scipdf_parser](https://github.com/titipata/scipdf_parser) | ~0.5k | Python parser for research PDFs (body + figures, based on GROBID) |
---
## 🕸️ Citation Graph and API Clients
> For citation network analysis or directly querying academic databases with code.
| Project | Stars | Description |
|---|---|---|
| [scholarly](https://github.com/scholarly-python-package/scholarly) | ~1.9k | Pythonic Google Scholar author/paper search |
| [semanticscholar](https://github.com/danielnsilva/semanticscholar) | ~0.5k | Unofficial Python client for Semantic Scholar API |
| [pyalex](https://github.com/J535D165/pyalex) | ~0.4k | Lightweight OpenAlex API Python interface |
| [ArxivDigest](https://github.com/AutoLLM/ArxivDigest) | ~0.5k | Personalized daily arXiv abstracts with GPT relevance scoring + email push |
| [citegraph](https://github.com/Citegraph/citegraph) | ~22 | Open-source visualization of 5M+ paper/citation network (CS literature) |
---
## ✍️ Paper Writing and Peer Review Assistants
> Drafting, polishing, and "AI pre-review" before submission.
| Project | Stars | Description |
|---|---|---|
| [lmms-lab-writer](https://github.com/EvolvingLMMs-Lab/lmms-lab-writer) | ~0.3k | Local-first agentic LaTeX writing for AI-assisted academic writing |
| [academic-writing-agents](https://github.com/andrehuang/academic-writing-agents) | ~0.1k | Claude Code plugin: 10+ expert agents collaborate on academic writing review, research, drafting, and polishing |
| [ai-peer-review](https://github.com/poldrack/ai-peer-review) | ~0.2k | Multi-LLM meta-review: independent reviews + synthesized meta-review |
| [open_reviewer](https://github.com/maxidl/openreviewer) | ~14 | Generate high-quality peer reviews for ML/AI conference papers, for pre-submission feedback |
| [academic-research-plugin](https://github.com/JeanDiable/academic-research-plugin) | ~17 | Claude Code plugin: literature review, paper review, citation management, searching arXiv/S2/DBLP, and identifying research gaps |
---
## 📖 Awesome Lists
> For a systematic understanding, start with these community-maintained lists.
| List | Stars | Description |
|---|---|---|
| [Awesome-LLM-Scientific-Discovery](https://github.com/HKUST-KnowComp/Awesome-LLM-Scientific-Discovery) | ~0.4k | EMNLP 2025 survey list: LLMs for scientific discovery |
| [Awesome-Auto-Research-Tools](https://github.com/handsome-rich/Awesome-Auto-Research-Tools) | ~1.1k | Automated literature search, paper reading, experiment management, and code generation |
| [awesome-ai-auto-research](https://github.com/worldbench/awesome-ai-auto-research) | ~0.4k | AI-automated research survey |
| [LLM4SR](https://github.com/du-nlp-lab/LLM4SR) | ~0.1k | LLMs for scientific research survey papers and resources |
| [awesome-ai-research-tools](https://github.com/0x11c11e/awesome-ai-research-tools) | ~55 | AI research tools for literature review, literature management, data analysis, etc. |
| [awesome-evidence-synthesis](https://github.com/evidencesynthesis-tools/awesome-evidence-synthesis) | ~19 | Open-source tools for systematic reviews, meta-analysis, and evidence synthesis |
---
## 🏢 Commercial / Closed-Source Tools (Reference)
> Not open-source but widely used in literature work, listed for reference.
- **[Elicit](https://elicit.com)** — Extract data, generate evidence tables from millions of papers
- **[Consensus](https://consensus.app)** — Semantic search engine for research questions
- **[Scite](https://scite.ai)** — Citation context analysis (support/oppose/mention)
- **[Undermind](https://undermind.ai)** · **[SciSpace](https://typeset.io)** — Deep literature search and reading assistant
- **[Research Rabbit](https://researchrabbit.ai)** · **[Connected Papers](https://connectedpapers.com)** — Citation relationship visualization exploration
---
## 🧭 How to Choose (Decision Table)
| Your Needs | Recommended |
|---|---|
| I use Claude Code and want an end-to-end "research → paper" workflow | **academic-research-skills** (domain-leading, most comprehensive) |
| I want a general-purpose "help me research this topic" autonomous agent | **GPT Researcher** / **STORM** |
| I want to generate wiki-style long papers with citations | **STORM / Co-STORM** |
| I want fully automated "from idea to publishable paper" | **AI-Scientist-v2** / **AutoResearchClaw** |
| I need to do citation-based Q&A on a pile of PDFs | **PaperQA / PaperQA2** |
| I'm doing a rigorous PRISMA systematic review (screening thousands of abstracts) | **ASReview** or **prismAId** |
| I want to convert PDFs into clean Markdown for LLM | **MinerU / Docling / marker** |
| I want to integrate literature capabilities into my AI client (MCP) | **paper-search-mcp / zotero-mcp** |
| I want to chat with my literature library directly in Zotero | **zotero-gpt / PapersGPT** |
| I want AI pre-review for my paper before submission | **open_reviewer / ai-peer-review** |
| I just want to browse the landscape | Above **Awesome Lists** |
---
## 📈 Star Trends
> Star history of representative projects in this area (auto-updated).
[](https://star-history.com/#Imbad0202/academic-research-skills&stanford-oval/storm&assafelovic/gpt-researcher&Future-House/paper-qa&SakanaAI/AI-Scientist&Date)
---
## 🤝 Contributing
Feel free to add missing excellent projects! Please read [CONTRIBUTING.md](CONTRIBUTING.md):
1. Fork this repository → Add a new row in the corresponding category
2. Keep the format consistent: `[Name](link) | Stars | One-sentence description`
3. **Update both English and Chinese READMEs** (`README.md` and `README.en.md`)
4. Only include **available, actively maintained, and strongly related to literature review** projects
5. Submit a PR and explain your recommendation in the description
You can also use the [issue template](.github/ISSUE_TEMPLATE/add-a-tool.yml) to recommend a tool with one click.
If you find this useful, please star it to help more researchers discover it!
## 📄 License
[](LICENSE)
This list is under [CC0-1.0](LICENSE) (public domain contribution). Each project's copyright belongs to its respective authors.
Connection Info
You Might Also Like
No recommendations available