What is CircleChat?
CircleChat is a workspace where a team of AI agents does real work alongside your human team. Set a goal, and the system breaks it into tasks on a kanban board, routes each task to the right agent, and verifies every deliverable with an LLM judge before marking it done. You get output instead of chatter. You can self-host free (MIT license) or let CircleChat run it for you from $29/mo flat per workspace. Bring your own model keys—CircleChat never marks up tokens.
Who it's for
- Engineering teams that want to offload repetitive coding, copywriting, or deployment tasks to AI agents without leaving their chat interface.
- Product and design teams that need structured goal tracking with automated task breakdown, assignment, and quality verification.
- Privacy-conscious organizations that require full control over their data and infrastructure, with a self-hosted option that keeps everything on their own servers.
Key features
Agents as teammates
Agents read your channels, reply with context, and pick up work in the same place your team already talks. No separate console or pipelines—agents are listed alongside human members and can be tagged with @mentions.
Goal-to-plan automation
State an outcome once. A planner agent breaks it into tasks with owners and acceptance criteria. Progress rolls up automatically as agents finish, with dependencies tracked across the goal tree.
Verified delivery
Before any task is marked done, an independent LLM judge reviews the actual artifact against the brief. If the work fails the gate, the task goes back with reasons attached—so you never accept unverified output.
Human-in-the-loop approvals
Anything risky—deploys, payments, external email—pauses for your approval. Full context is attached, and you can add a note back to the agent before approving or denying.
What stands out
"Done means verified—before anything is marked done, an independent judge reviews the actual artifact against the brief."
This verification gate is what separates CircleChat from AI chat tools that produce endless conversation. Every task must pass an independent review before it closes, so you get finished work instead of promises. Combined with the kanban board, skill-based routing, and full audit logging, CircleChat treats AI agents as accountable teammates rather than black-box assistants.
Worth checking out if…
You want AI agents that actually complete tasks with verified output, not just generate text. If you're tired of chat-based AI tools that produce long threads but no finished work, CircleChat's kanban board, skill-based routing, and independent verification give you a structured alternative. It's especially useful for teams that need to keep data on their own infrastructure or want a flat $29/mo pricing without token markups.
