Stop prompting. Start directing.
Now in Early AccessBring any agent, use any model, swap mid-session — no lock-in. One unified workspace for you to plan and delegate to agents that design and build. Pick up the thread when it matters, and build as fast as you can think.
Recursive is available to Supporters
Works with the agents you already use
Don't see yours? Ask Recursive to build a Plugin for it →
Everything you need at a glance. Plans, file edits, turns, artifacts — every session captures the full story so you never lose the thread. Model and provider agnostic: start in Codex, continue in Claude Code, or let an automation run it. No lock-in.
A full task board shared between you and your agents. Prioritize, break down, and delegate — or do the work yourself. Agents pick up tasks, report progress, and hand back results. You decide how much to drive.
Agents post here when they need you. Review decisions, approve changes, or jump in and take over.
Kick off any kind of work on a schedule. Code reviews at dawn, dependency audits every Friday, post-merge cleanup on every push. Set it once, agents handle the rest — even when you're not at your desk.
Access your workspace from anywhere — your laptop, your phone, or another machine entirely. Agents keep working while you're away, and you can check in, review progress, or take over from any device.
Everything lives as files on your machine — just like Obsidian. Agents get full context, and you can browse artifacts, plans, and memory directly in your file system. Extend with MCP servers, skills, rules, and integrations.
Build your own PluginCommon questions
What is Recursive?
Recursive is an autonomous development platform — think of it as the IDE layer between you and your AI agents. You give agents structured context (tasks, workflows, memory) and they do the work. You review the results and get things over the finish line.
How is this different from Cursor or Windsurf?
Recursive isn't built like a traditional editor — it's a platform layer that sits above your existing CLI and API agents and gives them the context they need to ship work autonomously, while giving you a familiar UI to review results and get things over the finish line. Bring any agent, extend it however you want.
Does it just push code without my approval?
No. Every change happens in an isolated worktree on a separate branch. Agents post to your inbox when they need decisions. Nothing touches your main branch until you explicitly review and merge.
What kind of tasks can it handle?
Features, bug fixes, refactors, test coverage, documentation — anything you'd put in a ticket. Agents work best with well-scoped tasks, but can break down larger goals into subtasks automatically.
Is it locked to a specific AI model or editor?
No. Sessions are adapter and model agnostic — start work in Claude Code, continue in Codex, and finish in Gemini, all with the same context. Bring your own API keys, swap providers mid-run, and set up fallbacks so work never stalls because one model is down.
Where does my data live?
Everything lives as files on your machine — sessions, plans, tasks, memory, artifacts. There's no proprietary database or cloud-only storage. You can browse everything in your file system, back it up however you want, and move it between machines. Think Obsidian, but for your development workspace.
Can I manage multiple projects?
Yes. Recursive gives you a unified workspace across all your repos. Jump between projects, see tasks across everything you're working on, and let agents work in parallel across different codebases. Each project keeps its own memory, skills, and configuration.
Can I access my workspace remotely?
Remote access is coming soon as an optional subscription add-on. You'll be able to check in from any device — your laptop, phone, or another machine — review progress, approve decisions, or take over from wherever you are while agents keep working.
How do I get access?
Recursive is currently available to supporters. See the pricing page for details.