# ai4s-research/open-science

Open Science Desktop — local-first, model-agnostic AI research workbench for macOS, Windows & Linux. Open-source Claude Science desktop alternative built on Tauri + MCP + agent skills.

Repository: https://github.com/ai4s-research/open-science
Canonical: https://ross.abutalabs.com/products/ai4s-research-open-science
Homepage: https://openedscience.com/
Language: TypeScript
License: NOASSERTION
License Family: other
Topics: ai-agent, ai-for-science, ai-scientist, ai4s, claude-science, desktop-app, local-first, mcp, open-science, reproducible-research, research-tools, scientific-research, tauri, claude-science-alternative, research-workbench, open-science-desktop, claude-science-desktop-alternative, macos, windows, linux
Last push: 2026-08-26T14:58:31+00:00

## Health v2 (maintenance only)
Score: 80/100 (v2, computed 2026-09-03T02:39:23.370411+00:00)
- activity 99, release rhythm 98, longevity 4
- inputs: {"age_days": 61, "days_push": 7, "days_rel": 15, "gap_med": 1, "n_releases_24m": 24}
- flags: young, no_license
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1491, forks 158 (observed 2026-08-28T04:04:52.565209+00:00)

## What it is
Open Science Desktop is a local-first, model-agnostic AI research workbench for macOS, Windows, and Linux, built with Tauri, MCP, and agent skills. It connects agents, notebooks, files, figures, reports, and runs into one auditable desktop workflow with full artifact provenance and approval-by-default execution.

## Use cases
- run AI-assisted scientific analysis with reproducible provenance
- search literature and draft manuscripts with citations
- run Python notebooks locally and keep the scripts behind every result
- drive AI agents from a desktop app with MCP connectors
- run a headless AI research workbench on a server via CLI
- audit how a figure or report was generated
- use local models instead of a cloud AI subscription

## When to choose
- you want an open-source alternative to Claude Science for AI-assisted research
- you need traceable, reproducible outputs linking figures to code and data
- you prefer local-first execution with explicit approval for risky actions
- you need a model-agnostic workbench supporting multiple agents and protocols

## When to avoid
- you need a fully signed, polished commercial installer
- you want a purely cloud-hosted collaborative research platform
- you need mobile or web-only access

## Facets
- artifact type: application
- maturity: active
- function: agent-framework, mcp, chat-interface, developer-tools, workflow-automation, data-science
- domain: artificial-intelligence, data-science, developer-tools, self-hosted, cross-platform
- platform: windows, cli
- tags: ai-research-workbench, local-first, tauri, reproducible-research, open-science, model-agnostic, provenance, scientific-research, agent-skills, acp, ai-agents, macos, linux, desktop

## Member repositories
- ai4s-research/open-science (main) score 80

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:04:52.565209+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T04:33:33.466913+00:00, confidence not recorded.
  - readme: https://github.com/ai4s-research/open-science (fetched 2026-08-28T04:04:52.565209+00:00, sha cacb20475c39)
  - homepage: https://openedscience.com/ (fetched 2026-08-29T11:39:41.054447+00:00, sha e9e312880c46)
  - site_page: https://openedscience.com/docs (fetched 2026-08-29T11:39:41.064071+00:00, sha 1b053503fd48)
- Data as of 2026-08-30T08:39:29.467469+00:00.