One agent to talk to.
A team to do the work.
Rig lets the coding agents you already use work together: one leads, the others help, everything is checked before it lands, and you can undo anything.
Claude Code
Codex
Cursor
opencode
Grok
pi
What you get on day one
When one plan runs out, work moves to another agent you already pay for.
Queue work before bed. Wake up to a brief with Undo on everything.
Tag an issue in GitHub, Linear, Jira, Asana or Notion. Get a draft PR.
Your teammates’ agents coordinate, so nobody edits the same file at once.
Change an API once; helpers update every repo that calls it.
How early access works
Takes ten seconds.
Which agents, which trackers, how big your team is. We invite people whose setup the current build handles best.
New waves every two weeks. Invite friends to move up the list.
Questions
The first wave opens in December. New waves open every two weeks after that.
No. Rig uses the agents you already have. If you have none, Rig’s own agent runs on credits.
It stays on your computer. Rig has no servers that hold your code or your agent sign-ins.
Yes, the core is on GitHub today. Early access is for the desktop and terminal apps.