OpenViSTAI · Agents · Blockchain · Crypto
Submit a project
Menu
Research Agents

Openscience

OpenScience is an open-source AI research workbench, available as a desktop app, browser workspace and CLI. It takes a research goal, plans work, gathers literature, runs code and experiments, and produces findings with a record of its steps.

Explore the practical guide ↓

Independent OpenViST project profile, based on linked sources.

Openscience — OpenViST project profile in Research Agents.
OpenViST project illustration. See the linked sources for product details.
Understand the project

Practical overview

When it fits

Use OpenScience when a research task involves literature, project files, analysis code or experiments and you want the work recorded for review. Start with a plan when the method should be agreed before execution.

Who it is for

  • Researchers who want help reviewing scientific literature and saving cited evidence.
  • People analyzing data, reproducing claims, or running experiments with reviewable code and outputs.
  • Labs that want to adapt research procedures or extend an agent with custom tools and workflows.

Documented uses

  • Search scientific sources, compare findings and save cited evidence.
  • Inspect data, run analysis code, and produce figures or reports.
  • Reproduce a claim by agreeing on prerequisites and budget, then comparing measured results.
  • Delegate bounded tasks to worker agents and review the lead agent's synthesis.
Before you choose

Requirements & limits

The project documents a desktop app for macOS, Windows and Linux, as well as CLI and browser-workspace use. The CLI can be installed with npm or run through npx. Ace is described as a managed pay-as-you-go model option.

  • The recorded session exposes searches, code runs and written files for review.
  • File access uses explicit read and write grants.
  • Local Python and R kernels are available; Modal, SSH, Slurm and PBS are listed for remote compute, with jobs approved before running.
  • Model access can use user-provided API keys, a supported provider sign-in, a local endpoint or Ace.
  • BioNeMo NIM adapters are experimental and require a user-supplied NVIDIA API key and approval for each dispatch.

Check current requirements in the official documentation. Source-based guidance does not mean OpenViST has run or independently tested this project.

Source-based practical guide

Start a research task in a project folder

Give OpenScience a project folder and a bounded research or analysis request.

Before you start

  • The OpenScience CLI is available.
  • A project folder containing the files needed for the task.

Steps

  1. Open the project by running openscience ~/research/my-project with the path to your folder.

  2. Describe the task in plain language, including requested outputs and any constraints such as keeping original data unchanged.

  3. Use /plan first when you want to agree on a method before work begins.

  4. Review the recorded sources, assumptions, code and outputs before relying on the result.

Expected outcome

The agent works through the requested task in the project and records its steps and outputs for review.

Guide limitations

  • A skill does not guarantee that every tool or service it references is installed; check availability in Customize.
  • Review the work and scientific conclusions rather than treating the recorded trace as independent validation.
Documentation 1 ↗
Evidence and transparency

Sources & review

Official repository ↗
Source record details
Source availability
Active
Sources checked
7 October 2026
Original website title
synthetic-sciences/openscience · Official homepage title · retained in its source language
How we review sourcesSuggest a correction