edict

Why we included this project

Teams tired of letting agents chat freely and hoping for the best will find a different approach here. Edict arranges a dozen specialized AI agents into a fixed pipeline borrowed from imperial Chinese governance: a dispatcher sorts incoming messages, a planner breaks commands into subtasks, a dedicated reviewer audits the plan and can reject it outright, and executor agents run the approved work in parallel before reporting back. That review step is the real difference, since it forces a quality check between planning and execution that most orchestration frameworks skip. A real-time dashboard tracks every agent, lets you stop, cancel, or resume tasks mid-flight, and keeps a full audit trail. The system runs on OpenClaw.