Skip to content
Back to search

CoAutoResearch

An open-source research system for autonomous investigation, recursive self-improvement, and human–AI collaboration. Run experiments, refine methods through evidence and feedback, and turn traceable research into paper drafts.

Safety scan

A static text scan, not a guarantee. Select the badge to see what was scanned and every finding.

Checked today
Badge for your README
findsafeskills safety rating

Markdown

[![findsafeskills safety rating](/api/listings/yihongt--coautoresearch/badge.svg)](/listings/yihongt--coautoresearch)

HTML

<a href="/listings/yihongt--coautoresearch"><img src="/api/listings/yihongt--coautoresearch/badge.svg" alt="findsafeskills safety rating" /></a>

Install

i

Install instructions from the README

As written by the repo's author. findsafeskills has not verified these steps.

**Supported coding agents:** ✓ Codex CLI · ✓ Claude Code

Give your coding agent this instruction:

Set up and launch CoAutoResearch from
https://github.com/YihongT/CoAutoResearch.
Follow docs/agent-setup.md, reuse my existing coding-agent login,
configure paper generation, verify the setup, and open the dashboard.

Setup has three checkpoints: **dashboard opens → coding agent is authenticated and ready → paper tools are ready**. Complete interactive provider login yourself. Model availability and usage limits depend on your account; research can require additional dependencies.

<details> <summary>Manual setup from source</summary>

You need Node.js 20+, Python 3.10+, Git, and an installed, authenticated Codex or Claude Code CLI.

git clone https://github.com/YihongT/CoAutoResearch.git
cd CoAutoResearch
node bin/auto-research.js doctor
node bin/auto-research.js ui

Open the printed URL and keep the server running. The dashboard needs no application dependencies or build step. Use `--projects-dir /path/to/projects` to keep research outside the checkout.

PDF generation additionally needs Python 3.12+, four pinned scientific skills, LaTeX and Poppler. Follow the agent setup guide. You can use this source checkout without an npm release.

[Agent setup](docs/agent-setup.md)

</details>

Your first session: **Create project → send a research brief → review the direction → Start autoresearch**. Include your materials, compute budget and decisions that require approval. Creating a project alone does not start research.

[Setup](https://yihongt.github.io/CoAutoResearch/agent-setup.html) · [Walkthrough](https://yihongt.github.io/CoAutoResearch/walkthrough.html)

<a id="how-it-works"></a>

What was scanned

The files the safety scan read, as written by the repository's author and shown as plain text. Lines with a finding are marked.

README.md · README · 189 lines

View this file on GitHub (read at the repository's current version: the scan did not record which commit it read)

  1. <!-- Translation basis is tracked in readme/translations.json. -->
  2. <p align="center"><picture>
  3. <source media="(prefers-color-scheme: dark)" srcset="assets/readme-hero-dark.svg">
  4. <img src="assets/readme-hero-light.svg" width="1200" alt="CoAutoResearch — autonomous research and human–AI collaboration">
  5. </picture></p>
  6. <p align="center"><strong>An autonomous research partner that self-improves recursively and works with you.</strong></p>
  7. <p align="center">Define a question. Run experiments. Discuss findings. Refine the next step.<br>Build a manuscript and paper draft from evidence you can trace.</p>
  8. <p align="center"><a href="#quick-start"><strong>Quick start →</strong></a> · <a href="#features">Features</a> · <a href="#example-papers">Example papers</a> · <a href="https://yihongt.github.io/CoAutoResearch/">Documentation</a></p>
  9. <details>
  10. <summary>🌐 Languages</summary>
  11. <p align="center"><strong>English</strong> · <a href="readme/README.zh-CN.md">简体中文</a> · <a href="readme/README.zh-TW.md">繁體中文</a> · <a href="readme/README.ja.md">日本語</a> · <a href="readme/README.ko.md">한국어</a> · <a href="readme/README.es.md">Español</a><br><a href="readme/README.pt-BR.md">Português brasileiro</a> · <a href="readme/README.fr.md">Français</a> · <a href="readme/README.de.md">Deutsch</a> · <a href="readme/README.ru.md">Русский</a> · <a href="readme/README.ar.md">العربية</a></p>
  12. <p align="center"><sub>These are README translations. The interface and full documentation are in English; agent replies follow your language.</sub></p>
  13. </details>
  14. <p align="center"><a href="LICENSE"><img src="https://img.shields.io/badge/license-Apache_2.0-d7b775" alt="Apache 2.0"></a> <a href="https://yihongt.github.io/CoAutoResearch/"><img src="https://img.shields.io/badge/docs-get_started-737269" alt="Documentation"></a></p>
  15. <a id="news"></a>
  16. ## 📰 News
  17. - **2026-09-11** — [CoAutoResearch v2.0.0](https://github.com/YihongT/CoAutoResearch/releases/tag/v2.0.0) released.
  18. <a id="features"></a>
  19. ## ✨ Features
  20. **Discuss while research runs.** Explore ideas in a separate chat. Review and send a suggestion when you want it to guide the research.
  21. **Recursive self-improvement.** Feed recorded results, review feedback and reusable lessons into the next trial. Refine hypotheses, methods and research decisions within your brief and budget; keep negative findings and stop when the evidence calls for it.
  22. **Inspect the evidence.** Follow findings, checks and limitations into a manuscript and a paper draft.
  23. <a href="https://yihongt.github.io/CoAutoResearch/walkthrough.html#screenshots">
  24. <picture>
  25. <source media="(prefers-reduced-motion: reduce)" srcset="assets/browser-tour.png">
  26. <img src="assets/browser-tour.gif" alt="Digits: Trial → Manuscript → Paper">
  27. </picture>
  28. </a>
  29. Browse a completed Digits project: research records, manuscript and PDF. This edited, cropped browser tour shows existing history, not a new run or actual research speed.
  30. <details>
  31. <summary>Three ways to begin</summary>
  32. **A new question:** “Compare two methods on a small public dataset. Use CPU only, state the budget, and ask before final evaluation.”
  33. **Existing work:** “Read my proposal and attached materials. Identify the next useful experiment and prepare a plan for review.”
  34. **A result to investigate:** “Check whether this finding survives a stronger comparison. Preserve the original results and report limitations.”
  35. </details>
  36. <a id="quick-start"></a>
  37. ## 🚀 Quick start
  38. **Supported coding agents:** ✓ Codex CLI · ✓ Claude Code
  39. Give your coding agent this instruction:
  40. ```text
  41. Set up and launch CoAutoResearch from
  42. https://github.com/YihongT/CoAutoResearch.
  43. Follow docs/agent-setup.md, reuse my existing coding-agent login,
  44. configure paper generation, verify the setup, and open the dashboard.
  45. ```
  46. Setup has three checkpoints: **dashboard opens → coding agent is authenticated and ready → paper tools are ready**. Complete interactive provider login yourself. Model availability and usage limits depend on your account; research can require additional dependencies.
  47. <details>
  48. <summary>Manual setup from source</summary>
  49. You need Node.js 20+, Python 3.10+, Git, and an installed, authenticated Codex or Claude Code CLI.
  50. ```bash
  51. git clone https://github.com/YihongT/CoAutoResearch.git
  52. cd CoAutoResearch
  53. node bin/auto-research.js doctor
  54. node bin/auto-research.js ui
  55. ```
  56. Open the printed URL and keep the server running. The dashboard needs no application dependencies or build step. Use `--projects-dir /path/to/projects` to keep research outside the checkout.
  57. PDF generation additionally needs Python 3.12+, four pinned scientific skills, LaTeX and Poppler. Follow the agent setup guide. You can use this source checkout without an npm release.
  58. [Agent setup](docs/agent-setup.md)
  59. </details>
  60. Your first session: **Create project → send a research brief → review the direction → Start autoresearch**. Include your materials, compute budget and decisions that require approval. Creating a project alone does not start research.
  61. [Setup](https://yihongt.github.io/CoAutoResearch/agent-setup.html) · [Walkthrough](https://yihongt.github.io/CoAutoResearch/walkthrough.html)
  62. <a id="how-it-works"></a>
  63. ## 🔄 How it works
  64. Each **Trial** is a bounded research iteration. The service checks proposed changes, records accepted results and determines whether to continue, pause or request a human decision. Internal review is not external peer review or proof of a scientific claim.
  65. **Add to research draft** prepares text; it does not send or apply it. **Pause after current turn** requests a pause at an agent-turn boundary, possibly inside a Trial. **Resume autoresearch** continues from the retained state.
  66. <picture>
  67. <source media="(prefers-reduced-motion: reduce)" srcset="assets/co-auto-light.svg">
  68. <source media="(prefers-color-scheme: dark)" srcset="assets/co-auto-dark.gif">
  69. <img src="assets/co-auto-light.gif" alt="How it works">
  70. </picture>
  71. Illustrated workflow; timing is schematic. Static alternatives are available. [SVG](assets/co-auto-light.svg)
  72. <details open>
  73. <summary>🏗️ Explore the system architecture</summary>
  74. <picture>
  75. <source media="(prefers-reduced-motion: reduce)" srcset="docs/_static/diagrams/architecture.svg">
  76. <source media="(prefers-color-scheme: dark)" srcset="docs/_static/diagrams/architecture-dark.gif">
  77. <img src="docs/_static/diagrams/architecture.gif" alt="Explore the system architecture">
  78. </picture>
  79. [Architecture and lifecycle](https://yihongt.github.io/CoAutoResearch/conceptual-framework.html)
  80. </details>
  81. <a id="example-papers"></a>
  82. ## 📄 Example papers
  83. <table>
  84. <tr>
  85. <td width="50%" valign="top"><a href="docs/_static/examples/digits-paper.pdf"><img src="docs/_static/examples/digits-preview.png" width="360" alt="Can covariance shrinkage help classify handwritten digits with little training data?"></a><p><strong>Can covariance shrinkage help classify handwritten digits with little training data?</strong></p><p>Matched small-sample comparisons and a single held-out evaluation; findings remain specific to this dataset and protocol.</p><p><a href="docs/_static/examples/digits-paper.pdf">Read paper · PDF</a></p></td>
  86. <td width="50%" valign="top"><a href="docs/_static/examples/ising-paper.pdf"><img src="docs/_static/examples/ising-preview.png" width="360" alt="How reliably can a laptop estimate an Ising phase transition?"></a><p><strong>How reliably can a laptop estimate an Ising phase transition?</strong></p><p>CPU-bounded estimation and implementation diagnosis. Partial gains did not meet the joint criteria; held-out seeds stayed unused.</p><p><a href="docs/_static/examples/ising-paper.pdf">Read paper · PDF</a></p></td>
  87. </tr>
  88. </table>
  89. Both papers are developer-operated research drafts with human intervention, not external user case studies. Their limitations remain in the reports; human scientific review is required before publication.
  90. When reviewed results are recorded and project agents are idle, choose **Generate paper**. The internal agent uses writing, visualization, citation and venue skills on a frozen evidence snapshot, without running new experiments.
  91. Open **Paper** to zoom, scroll and download the PDF and source. A previous draft remains available while its replacement is generated.
  92. [Example papers](https://yihongt.github.io/CoAutoResearch/example-papers.html) · [Paper generation](https://yihongt.github.io/CoAutoResearch/paper-generation.html)
  93. <a id="documentation"></a>
  94. ## 📚 Documentation
  95. The linked guides are in English. Start with Setup or Walkthrough; consult FAQ for common questions and Platforms for support boundaries.
  96. - [Setup](https://yihongt.github.io/CoAutoResearch/agent-setup.html)
  97. - [Walkthrough](https://yihongt.github.io/CoAutoResearch/walkthrough.html)
  98. - [Paper generation](https://yihongt.github.io/CoAutoResearch/paper-generation.html)
  99. - [FAQ](https://yihongt.github.io/CoAutoResearch/faq.html)
  100. - [Platforms](https://yihongt.github.io/CoAutoResearch/platforms.html)
  101. - [CLI](https://yihongt.github.io/CoAutoResearch/cli.html)
  102. - [Architecture](https://yihongt.github.io/CoAutoResearch/conceptual-framework.html)
  103. - [Upgrades](https://yihongt.github.io/CoAutoResearch/upgrading.html)
  104. <a id="community"></a>
  105. ## 🤝 Community
  106. Help verify installation on your platform, report reproducible bugs, improve research workflows, maintain translations, or share documented research outcomes. Use Issues and read the contribution guide; report security issues privately.
  107. [Issues](https://github.com/YihongT/CoAutoResearch/issues) · [Contributing](CONTRIBUTING.md) · [Translations](readme/TRANSLATING.md) · [Security](SECURITY.md) · [Code of conduct](CODE_OF_CONDUCT.md)
  108. <details>
  109. <summary>Cite CoAutoResearch</summary>
  110. Cite the software and record the version or commit used.
  111. ```bibtex
  112. @software{coautoresearch,
  113. title = {CoAutoResearch},
  114. author = {CoAutoResearch contributors},
  115. url = {https://github.com/YihongT/CoAutoResearch}
  116. }
  117. ```
  118. </details>
  119. Licensed under Apache 2.0. Upstream tools and skills retain their own licenses and attribution requirements.
  120. Contact: [yihong.tang@mail.mcgill.ca](mailto:yihong.tang@mail.mcgill.ca)

Ratings and comments

Sign in to rate this skill.

Sign in

Free to use. If it’s helped, pass it on.