Practical overview
When it fits
Use it when a scientific project benefits from coordinating research requests, files, scientific data access and computational work in one desktop workspace, particularly when reviewable outputs and traceable artifacts matter.
Who it is for
- Scientists and researchers working on computational or data-intensive projects.
- Research teams that need to review, hand off or archive agent-assisted research sessions.
- Users seeking a desktop workspace that combines scientific agents with Python or R workflows.
Documented uses
- Turn a research question, files and expected outputs into an agent-assisted project task.
- Search literature or scientific data sources and develop research hypotheses.
- Run code, analyze data, simulate or visualize results, and create reports, tables or figures.
- Review tool activity and provenance associated with generated artifacts.
- Package a completed session for review, handoff or archiving.
Requirements & limits
The repository describes the application as open source under the Apache-2.0 license. Desktop installers are documented for macOS 12+, Windows x64 and Linux x64; macOS installation through Homebrew is also described.
- The application is local-first and model-agnostic; setup includes configuring a model provider and validating the connection.
- Python/R Notebook setup is optional, while the required environment and agent-runtime checks are part of first-time setup.
- Research tasks can use attached files, skills, MCP tools and scientific connectors; available workflows depend on the configured components.
- Provenance provides a place to inspect available evidence for generated artifacts.
- Portable .science packages can include selected conversation branches, file versions, Notebook records and verification evidence.
Check current requirements in the official documentation. Source-based guidance does not mean OpenViST has run or independently tested this project.
Create a research project and review its outputs
Submit a research task in a project and inspect the resulting artifacts and available provenance.
Before you start
- Complete the app's required first-time setup, including a tested model connection.
- Have the research goal, inputs, and expected outputs ready; files to attach are optional.
Steps
Click **New project**, open a session, and describe the research goal, inputs, and expected outputs.
Attach files, select a model and approval mode, then send the task. Use
@to reference project files or/to choose a skill.Review tool activity and approval requests, preview the results, and check available evidence in **Provenance**.
Expected outcome
The task is submitted in a project session, with results available for preview and any available provenance evidence available for review.
Guide limitations
- The README notes that screenshots illustrate the workflow and that interface labels, catalogs, and other details may differ by installed version.
- The guidance does not specify particular models, skills, or task-specific validation criteria.
Sources & review
Source record details
- Source availability
- Active
- Sources checked
- 9 October 2026
- Original website title
- aipoch/open-science · Official homepage title · retained in its source language