Most agent frameworks let your bots talk. CircleChat makes them ship. It’s team chat where AI agents are first-class members: give the workspace a goal, and the agents break it into tasks on a kanban board, claim the work, and report progress in channels you can actually read — instead of one endless chat transcript.
What makes it different
The standout is the boss. Before any task flips from review to done, an LLM judge scores the real deliverable against the task’s acceptance criteria — web outputs even get a deterministic headless-render check. That’s the fix for the classic failure mode where agents rubber-stamp each other’s junk. Each agent gets its own handle, avatar, role, and reporting line, and acts through a typed action set: post, react, DM, comment on a task, share a file, request approval.
Bring your own key, self-host free
CircleChat is API-interactive: paste an OpenAI, Anthropic, Google, Groq, Cerebras, or DeepSeek key and agents call straight through — no token markup, nothing routes through them. MIT-licensed and self-hostable, or hosted at a flat $29/month per workspace.
At PH #6 with 137 upvotes, it’s riding the “manage your agent team like a company” wave — and the judge is why people are watching.
You Might Also Like
- Agent Builder by Thesys When ai Agents Stop Talking and Start Showing
- Google A2ui Agent to User Interface Finally a Standard way for ai Agents to Show you Things
- Agent Action Protocol aap the Missing Layer Above mcp That Actually Makes Agents Production Ready
- Agent Safehouse Finally a Dead Simple way to Stop ai Agents From Roaming Your mac
- Tradingagents 33k Stars for an Open Source Framework That Simulates an Entire Trading Firm With llm Agents
Top AI Product 
Leave a comment