OpenResearch: A local-first workspace for research agents

AlphaXiv released OpenResearch, a local-first workspace that turns coding agents including Claude Code, Codex, OpenCode, Cursor, and Google Antigravity into research agents that can review literature, develop hypotheses, run experiments, and produce research artifacts. The tool installs via a CLI script on macOS and Linux (glibc 2.35+) and a Windows beta requiring Git for Windows, opening a local dashboard at http://127.0.0.1:4791, and gives each research direction an independent agent session with an isolated git worktree and an immutable archive of each run's recorded commit. The same committed source snapshot can run locally, over SSH, or on Slurm, Kubernetes, Ray, Hugging Face Jobs, Modal, Tinker, and managed OpenResearch compute, with projects, conversations, experiments, runs, logs, code, and artifacts kept on the user's machine.

The local-first workspace for research agents and autoresearch. Turn Claude Code, https://github.com/alphaXiv/OpenResearch/blob/main/.github/readme-assets/claude.svg Codex, https://github.com/alphaXiv/OpenResearch/blob/main/.github/readme-assets/codex.svg OpenCode, https://github.com/alphaXiv/OpenResearch/blob/main/.github/readme-assets/opencode.svg Cursor, or Google Antigravity into research agents that can review literature, develop hypotheses, run experiments, and produce research artifacts. https://github.com/alphaXiv/OpenResearch/blob/main/.github/readme-assets/cursor.svg <sub macOS 11+ · Windows beta requires Git for Windows https://github.com/alphaXiv/OpenResearch/blob/main/docs/windows.md · Linux app needs glibc 2.35+</sub Install the CLI on macOS or Linux, then launch OpenResearch: curl -LsSf https://openresearch.sh/install.sh | sh orx up orx up opens the local dashboard at http://127.0.0.1:4791 . On a managed Mac for example, a work computer , use the macOS download above instead. Device-management policies may block the CLI that install.sh installs because it is not yet signed. The app is signed with a Developer ID and notarized by Apple. To use orx in your terminal, click Install under Install the orx command in the app's Settings → Updates, or run adjusting the path if the app is not in /Applications : /Applications/OpenResearch.app/Contents/MacOS/orx install-cli Either way, orx is linked into ~/.local/bin , with a hint to add it to your PATH if needed. If you already ran install.sh , remove ~/.cargo/bin/orx first and open a new terminal. On Windows, use the beta download above after installing Git for Windows https://github.com/alphaXiv/OpenResearch/blob/main/docs/windows.md . It installs for your account, with no administrator prompt. The installer isn't signed yet, so Windows may say "Windows protected your PC": choose More info → Run anyway . On Linux, the desktop app is an AppImage for x86 64 https://github.com/alphaXiv/OpenResearch/releases/latest/download/OpenResearch-x86 64.AppImage or ARM64 https://github.com/alphaXiv/OpenResearch/releases/latest/download/OpenResearch-aarch64.AppImage . Keep it somewhere you can write to, such as ~/Applications , so it can update itself, then change to that directory, run chmod +x OpenResearch- .AppImage , and open it. See Linux https://github.com/alphaXiv/OpenResearch/blob/main/docs/linux.md for requirements. Connect a local model https://github.com/alphaXiv/OpenResearch/blob/main/docs/local-models.md to use LM Studio, oMLX, Ollama, or a custom endpoint with OpenCode. Create an account at openresearch.sh https://openresearch.sh to receive email updates and use managed OpenResearch compute. | | OpenResearch gives you | |---|---| | Parallel exploration | Give each research direction an independent agent session and isolated git worktree. | | Reproducible experiments | Track variants in a git-native experiment tree; every run receives an immutable archive of its recorded commit. | | Evidence in context | Keep logs, diffs, files, results, and artifacts tied to the work that produced them. | | Your choice of agent | Use Claude Code, Codex, OpenCode, Cursor, or Google Antigravity, with the harness and model selected per session. | | Your choice of compute | Run locally, on your own infrastructure, or with managed OpenResearch compute. | | Local ownership | Keep projects, conversations, experiments, runs, logs, code, and artifacts on your machine. | OpenResearch can run the full loop autonomously: propose an idea, change the code, launch an experiment, inspect the evidence, and decide what to try next. Multiple agents can explore different directions in parallel while the experiment tree preserves their lineage. The same committed source snapshot can run locally, over SSH, or on Slurm, Kubernetes, Ray, Hugging Face Jobs, Modal, Tinker, and managed OpenResearch compute. Publishing the repository is not required. Run the workspace next to remote GPUs while using the browser on your laptop: orx up --remote user@host SSH config aliases and custom ports are supported. The remote service binds to loopback and has no application-level authentication, so other users on that host can reach it. Install the OpenResearch skill into supported coding agents: orx install-skills Common commands: orx projects orx project view <project-id orx runs <project-id orx logs <run-id orx exp run <experiment-id orx discover keyword <query orx paper <arxiv-id-or-doi Run orx --help or orx <command --help for the complete interface. OpenResearch runs on 127.0.0.1 with a local SQLite store. Creating a project or launching a run does not publish your code. An openresearch.sh https://openresearch.sh account is only used for service-owned capabilities such as organizations and managed compute. Official release builds send opt-out, coarse usage events tied to a random installation ID. They do not include code, prompts, file contents or paths, repository names, tokens, emails, or project and experiment identifiers. orx telemetry off orx telemetry status orx <command --no-telemetry Source and development builds do not send analytics. Coding agents may also file product feedback with orx feedback when you hit a bug, wish for a feature, or get frustrated with OpenResearch. Each report is a short description of the workflow problem. Bug reports include as much detail as possible to reproduce a failure while omitting sensitive information. Like analytics, reports are sent only from official release builds. They are linked to your account when you are logged in and turned off by orx telemetry off . The --no-telemetry flag covers only the command it is passed to.