AgentIndex icon
AgentIndex
ToolsMCP ServersHOTCategoriesTrendingNewBest ListsCompareGuides
Submit Tool
ToolsMCP ServersHOTCategoriesTrendingNewBest ListsCompareGuides
Home/
Data Processing/
open-science
open-science logo

open-science

Active·★ 5.5k·Apache-2.0·Updated 2026-11-17

AIPOCH Open-Science is an open-source, local-first research workbench (Apache 2.0) engineered for scientists, computational researchers, and data practitioners across machine learning, life sciences, physics, and materials engineering. Designed around local data privacy and reproducible methodology, it provides a transparent alternative to opaque cloud AI research tools.

The workbench orchestrates autonomous scientific agents across multiple frameworks (supporting Pi and OpenCode), backed by isolated Python and R execution environments. Researchers can ingest academic papers directly via DOI or arXiv links, read and annotate split-screen PDFs with multi-lingual OCR, execute data analysis pipelines, and generate publication-ready tables and figures with traceable computational provenance.

A standout feature is the exportable '.science' bundle format, which encapsulates conversation threads, source code versions, Jupyter notebook sessions, and verification artifacts into a portable archive for peer review and collaboration. While exceptionally capable, the local desktop application demands significant local memory and compute when running local OCR or executing multi-step computational simulations.

This write-up is based on the project's README and public repository data. We have not hand-tested the tool. Data as of 2026-11-12.

License
Apache-2.0
Latest release
v1.2.1 · 2026-11-11
Last commit
2026-11-12
Contributors
30
Forks
555
Open issues and PRs
35
First published
2026-07-03
npm package
@aipoch/open-science 1.2.1 · 664 downloads last week

Source: GitHub, npm and PyPI public APIs.

#agent-framework#agent-skills#ai-agent#ai-for-science#ai-research#ai-research-workbench#ai-workbench#desktop-app
$ Install
$ npm install --global @aipoch/open-science
↗ Visit site★ GitHub
01

Features

01Scientific agents with Python and R execution
02Scientific data connectors and web search in one workspace
03Reports, tables and figures with traceable provenance
04Exports sessions as portable `.science` packages with verification evidence
05Adds references from a DOI or arXiv link
02

Why choose it

+Strictly local-first design guarantees sensitive experimental datasets and unpublished manuscripts remain private.
+Sandboxed Python and R code execution sandboxes ensure verifiable mathematical and computational results.
+Unique .science package encapsulates full provenance, notebook checkpoints, and figures for peer review.
+Direct literature ingestion via DOI and arXiv alongside split-view PDF reader with multi-lingual OCR.
03

Trade-offs

!Heavy desktop footprint with high CPU and RAM demands when executing local OCR models or complex simulations.
!Quality of scientific hypothesis generation relies heavily on the reasoning capabilities of the configured host LLM.
!Extended research sessions with dense literature ingest can consume substantial API tokens if using commercial models.
!Bundled .science export packages can quickly grow to multiple gigabytes when including large scientific datasets.
04

Maintenance check

ActiveChecked: 2026-11-12
Last commit
2026-11-12
Commits, last 90 days
100+
Releases, last 90 days
30
Latest release
v1.2.1 (2026-11-11)
Open issues
36
Recent issues with at least one comment
0/7
Commits per week, last 12 weeks
Commits per week, last 12 weeks: 313, 293, 247, 274, 201, 143, 109, 161, 171, 113, 127, 114

Oldest week on the left; the latest four weeks are orange. Source: GitHub commit activity.

Status rule: archived on GitHub = Archived; last commit within 90 days = Active; 91 to 365 days = Slowing; older = Stale. All figures come from the GitHub API on the date shown.

Most discussed open issues

  • #3302 [Feedback]: Allow assigning new projects to custom folders for file reading and output generation1 comments · 2026-10-05
  • #925 [Feature]: RO-Crate export of artifact provenance1 comments · 2026-08-07
  • #4308 ci(windows-full-test.yml): scheduled run failing0 comments · 2026-11-12
  • #4302 ci(source-regression.yml): scheduled run failing0 comments · 2026-11-11
  • #2728 [Reproducibility Pilot] Review the first 10 cases and verify an improvement0 comments · 2026-09-17

Issue titles are quoted from GitHub. Open the link to check whether a problem still applies to the version you use.

05

Compatibility

macOS
Desktop app
Verified via docs
Windows
Desktop app
Verified via docs
Linux
Desktop app
Verified via docs
06

Quick start

1
$ npm install --global @aipoch/open-science
07

Use cases

↳Moving from literature review to hypotheses and data analysis
↳Producing reproducible analyses in life sciences, chemistry or physics
↳Handing off or archiving a finished research session
08

Alternatives

⚡Top VS:
VSeasy-stock★1.9k→VSdaily_stock_analysis★73.1k→VSgenoffice★12.5k→VSOpenBiliClaw★3.8k→
easy-stock logo
easy-stock★ 1.9k
A股行情分析与AI智能投研智能体:股票分析、量化交易分析、盘后复盘桌面工作台——easy stock
VSCompare→
daily_stock_analysis logo
daily_stock_analysis★ 73.1k
LLM 驱动的多市场股票智能分析系统:多源行情、实时新闻、决策看板与自动推送,支持零成本定时运行。 LLM-powered multi-market stock analysis system with multi-source market data, real-time news, decision dashboard, automated notifications, and cost-free scheduled runs.
VSCompare→
genoffice logo
genoffice★ 12.5k
Free, open-source AI Office suite: Docs, Sheets, Slides, PDF, Markdown and HTML editors with a built-in AI agent, plus a `genoffice` CLI and agent skill so Claude Code, Codex and Cursor can create and edit real .docx/.xlsx/.pptx files locally. Bring your own key. macOS, Windows & Linux.
VSCompare→
OpenBiliClaw logo
OpenBiliClaw★ 3.8k
本地私有、开源的自进化跨平台 AI 内容发现 Agent:先理解你,再主动从 B站、小红书、抖音、YouTube、X、知乎、Reddit、微博等平台与开放 Web 寻找内容。(支持 deepseek harness 插件) | Local-first open-source cross-platform AI content discovery agent: understands you, then proactively finds content across Bilibili, Xiaohongshu, Douyin, YouTube, X, Zhihu, Reddit, Weibo and the open web.(support deepseek harness plugin)
VSCompare→
autoclip logo
autoclip★ 10.7k
AutoClip|一个链接,一键出片。开源 AI 视频剪辑桌面工具,将播客、访谈、课程等长视频自动剪成短视频,生成字幕、封面和发布文案,适配抖音、小红书、TikTok、Reels 与 YouTube Shorts。Open-source AI video clipping & content repurposing.
VSCompare→
PI-Desktop logo
PI-Desktop★ 8.0k
Local-first AI coding agent desktop: Electron + Rust host core + pi Agent Harness + user-installable plugins
VSCompare→
oh-story-claudecode logo
oh-story-claudecode★ 8.7k
Claude Code / Codex / OpenCode agent skills for writing Chinese web novels (网文): 扫榜、拆文、写作、去AI味、封面全流程,长篇短篇都支持 | 14 skills for novel writing and long-form fiction with file-based continuity tracking. MIT.
VSCompare→
AnyModelBot logo
AnyModelBot★ 4.4k
🤖 Open-source, local-first alternative to Grok Bot, OpenAI dots and Claude Duet. Your own AI bots on any model: GPT, Claude, Gemini, Grok, DeepSeek, Kimi, GLM and 70+ more through one Modelflare key. Reach them in the app, WeChat or Feishu; they remember you, work in the background, run commands and use your computer. Pay per use.
VSCompare→
See all alternatives →

Comments

Log in to leave a comment

No comments yet. Be the first!

On this page
01Features02Why choose it03Trade-offs04Maintenance check05Compatibility06Quick start07Use cases08Alternatives
aipoch /
open-science
↗

Local-first, model-agnostic AI research workbench for reproducible computational science and literature review

5.5k555Apache-2.0
Last commit1d ago
StatusActive
CategoryData Processing
Contributors30
Latest releasev1.2.1
Links
Documentation↗Discussion↗Issues↗Releases↗
Cloud Infrastructure
Ad

Deploy on DigitalOcean

Host your MCP servers, Docker containers, and AI agents with $200 free credit over 60 days.

Claim $200 Free Credit
© 2026 AgentIndex.app|Built by a 10-year iOS Developer.
Best by use caseAboutContactAdvertisePrivacy PolicyTerms of UseQYSGitHubBuy me a coffee ☕

Browse by Category

MCP ServersCode AssistantWorkflow AutomationRAG / Knowledge BaseMulti-AgentBrowser AutomationLLM InfraDev ToolingObservability

Not affiliated with Anthropic, OpenAI or Microsoft.