# OpenResearch

A local-first workspace that turns coding agents such as Claude Code, Codex, OpenCode, Cursor or Google Antigravity into research agents that review literature, develop hypotheses, run experiments and produce research artifacts.

OpenResearch installs a CLI and launches a local dashboard with `orx up`, giving each research direction its own agent session and isolated Git worktree so several lines of inquiry can run in parallel. Experiments are tracked in a git-native tree, with every run receiving an immutable archive of the commit that produced it.

It keeps logs, diffs, files, results and artifacts tied to the work that produced them, and lets a user choose the agent and harness (Claude Code, Codex, OpenCode, Cursor or Google Antigravity) and the compute (local, a user's own infrastructure, or managed OpenResearch compute) per session. An 'Autoresearch' mode can run the loop autonomously: propose an idea, change the code, launch an experiment, inspect the evidence, and decide what to try next.

It is written in Rust, supports macOS 11+ and a Windows beta requiring Git for Windows, and is licensed under MIT.

Website: https://openresearch.sh/