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

Faros

FAROS is a Python-based, multi-agent research workflow system that connects literature search, experiment planning and execution, paper writing, and evidence review, with configurable model providers and human oversight.

Explore the practical guide ↓

Independent OpenViST project profile, based on linked sources.

Faros — 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 FAROS when research work needs to connect literature, an approved plan, executable experiments, paper drafts and evidence checks, with a researcher making decisions at key stages.

Who it is for

  • Researchers coordinating literature review, experiment development and paper preparation
  • Teams that want research plans, code, results and review records linked in a continuing workflow
  • Developers configuring a self-hosted AI research system with model providers and execution tools

Documented uses

  • Find and organize literature to develop evidence-supported research ideas
  • Turn a research plan into code, run experiments and retain configurations, logs and metrics
  • Compare results and generate figures for research analysis
  • Draft papers with citations and LaTeX/PDF output
  • Check claims against sources, metrics and figures, then review structured revision proposals
Before you choose

Requirements & limits

The repository is public and lists an MIT license. The documented quick start describes running a local backend and frontend.

  • The system is a research workflow runtime, not just a literature-search or paper-drafting tool; its scope also includes experiment implementation and evidence review.
  • A model service and its credentials must be configured for model calls; model capability requirements vary by task.
  • Generated-code execution requires separately configured Docker sandboxes, experiment dependencies and resource limits.
  • Paper PDF generation has additional LaTeX and CJK font requirements.

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

Set up FAROS and configure a model provider

Start the documented backend and frontend, then configure a model service for research workflows.

Before you start

  • Python 3.11 or later and Node.js 18 or later
  • Bash on Linux or WSL
  • A model service endpoint, model ID and credentials

Steps

  1. Clone the FAROS repository, create a Python virtual environment in the backend directory, install requirements.txt and start the FastAPI server with the documented Uvicorn command.

  2. In a second terminal, run npm ci in the frontend directory and start the frontend with VITE_API_BASE_URL set to the local backend address.

  3. In Settings / LLM Providers, enter the provider credentials, Base URL and model ID; save, test the connection and activate the provider and model.

Expected outcome

The documented local workbench is started and has an active model provider configured.

Guide limitations

  • These setup steps do not configure the Docker sandbox or experiment dependencies needed to run generated code.
  • LaTeX/PDF compilation requires additional tools and fonts.
Documentation 1 ↗
Evidence and transparency

Sources & review

Official repository ↗
Source record details
Source availability
Active
Sources checked
3 November 2026
Original website title
OpenNSWM-Lab/FAROS · Official homepage title · retained in its source language
How we review sourcesSuggest a correction