Woodor Plugins — The open-source toolkit for running AI agents at scale
Four plugins that work as one system: keep your AI spend under control, and keep your agents working together instead of in isolation. Free, open-source, installed in seconds.
/plugin marketplace add woodor-ai/plugins
* Compatible with Claude Code and Codex TUI.
/save-moneyLower AI bills, without changing how you work. It quietly steps in before you blow your budget, trims runaway output, and sends the cheap tasks to a cheaper model — running in the background so you don't have to think about it.
/init-agentsStop overpaying for simple work. This sets up your project so routine tasks go to a cheaper model and only the hard architectural calls go to the expensive one — automatically, with one command. You get the right answer without burning premium tokens on grunt work.
/handoffNever re-explain where you left off. When a session ends, it leaves a short note — what's done, what's pending, what to do next — and your next session picks up right where you stopped. No more copy-pasting context to get going again.
/agent-meetingYour AI agents stop working in isolation. Sessions running in different windows — or on different machines — can now message each other, chat as a group, and pull each other in to help, like coworkers in a room. (Under the hood: peer-to-peer discovery across your local network, no server to set up.)
Mission control for your agents — in your pocket
Your agents run across machines and sessions. The console puts the whole fleet on your phone — see what every agent is doing, keep spend in check, and step in from anywhere.
- See the whole fleet — every agent, every machine, one screen.
- Stay on budget — live spend, caps, and alerts before bills run away.
- Decide on the go — approve, pause, or redirect any agent.