{
  "id": 7138171,
  "title": "OpenResearch allows multiple research agents to run simultaneously without having to send out code.",
  "url": "https://urgent.news/2026/09/13/openresearch-agent",
  "topic": "tech",
  "section": "Tech",
  "published": "2026-09-13T17:28:05.000Z",
  "source": {
    "name": "Dev.to",
    "slug": "dev-to",
    "url": "https://dev.to/sarantoon/openresearch-aih-agent-wicchayhlaaythisthaangphrmkan-odyaimtngsngokhdkaip-525a"
  },
  "original_language": "th",
  "account": "OpenResearch, a tool from the alphaXiv team, allows multiple AI research agents to explore different directions simultaneously without requiring code sharing. The tool, written in Rust, provides features such as reproducible experiments, evidence tracking, and parallel exploration, making it easier to track the origin of results. OpenResearch also includes an Autoresearch mode that automates the entire research cycle, from proposing ideas to running experiments and deciding on next steps. The tool prioritizes data privacy, running locally with a SQLite database and providing options to opt-out of telemetry. It is available on GitHub with 1,863 stars and 135 forks as of September 13, 2026.",
  "summary": "OpenResearch is a project by alphaXiv that allows users to run parallel research agents with any model, without having to send code out. The project, written in Rust and licensed under MIT, was created on June 7, 2026, and has 1,863 stars with 135 commits and the latest push on September 12, 2026. The project's main features include reproducible experiments, evidence in context, parallel exploration, and autoresearch. Reproducible experiments allow users to track variants in the tree of experiments and archive immutable commits, while evidence in context stores logs, diffs, files, and outputs alongside the work that created them. Parallel exploration enables each research path to have its own agent session and git worktree, avoiding conflicts that arise from multiple agents editing the same file. Autoresearch runs the full research pipeline automatically, including proposing ideas, editing code, running experiments, and making decisions on what to try next. The project is designed to run locally on 127.0.0.1 with SQLite as the database and offers an opt-out telemetry feature.",
  "key_points": [],
  "editors_take": null,
  "illustration": null,
  "coverage": {
    "outlets": 1,
    "also_reported_by": []
  },
  "ai_generated": true,
  "disclaimer": "Summaries, key points and the editor’s take are written by software from other outlets’ reporting and may contain errors — always check the linked original."
}