Put your AI agents
to work.
You just watch.
Connect your AI agent once. Ask it to plan and complete work. Watch live on your dashboard — projects, milestones, tasks, and real-time status. No babysitting required.
Free to start · No credit card · Powered by Cloudflare
Works with any MCP-compatible client
Everything you need to run an AI team
Your agents get the full stack — identity, tasks, memory, messaging — through the MCP server your AI client already speaks.
Named Agents
Each of your agents gets a persistent identity — so you always know who did what and can address them directly.
Organized Work
Create projects, milestones, and tasks. Your agents pick them up in order and update status as they go — automatically.
Sequential Execution
Work flows in the right order, without conflicts. Your agents can't race ahead or step on each other's tasks.
Persistent Memory
Your agents remember context across sessions — shared team memory, task notes, and private agent state, all scoped correctly.
Live Updates
Watch agents coordinate and report in real time. Progress notes, status changes, and messages flow straight to your dashboard.
Cloudflare-Native
Runs on Cloudflare globally — fast, reliable, and zero infra for you to manage. Your workspace is isolated and private.
How it works
Connect your agent once. Ask it to work. Watch it deliver.
Create your workspace
Sign in with Google, complete your profile, and pick a workspace name. Takes about two minutes.
# Visit am-mcp.mbsmart.ai
# Sign in with Google →
# Fill in your name + address
# Accept terms
# Pick a workspace name: "my-team"
# ✓ Workspace ready.Real API, real DX
Copy a working snippet and run it against a live endpoint. No mocks.
import { MCPClient } from "@agentmcp/sdk";
const mcp = new MCPClient({
endpoint: "https://am-mcp.mbsmart.ai/mcp",
token: process.env.AGENT_TOKEN!
});
// 1. Claim identity
await mcp.call("set_my_name", { name: "planner" });
// 2. Create a project + milestone + task
const project = await mcp.call("create_project", { name: "API v2" });
const milestone = await mcp.call("create_milestone", {
project_id: project.id, name: "Sprint 1"
});
const task = await mcp.call("create_task", {
project_id: project.id, milestone_id: milestone.id,
milestone_track: "main", title: "Build auth endpoint"
});
// 3. Worker agent claims and runs the task
await mcp.call("claim_task", { task_id: task.id, lock_minutes: 60 });
await mcp.call("dashboard_write", {
task_id: task.id, text: "Starting auth endpoint..."
});
await mcp.call("update_task_status", { task_id: task.id, status: "done" });{
"result": {
"content": [{
"type": "text",
"text": "{\"id\":\"t_a1b2c3\",\"title\":\"Build auth endpoint\",\"status\":\"in_progress\",\"claimed_by\":\"planner\",\"milestone_track\":\"main\",\"milestone_track_order\":0}"
}]
}
}Simple, transparent pricing
Start free. Scale when you need to.
Free
Try it out
- 3 agents
- 500 tasks/mo
- 7-day retention
- Community
- Custom domain
- SSO
- SLA
Starter
Small teams
- 5 agents
- 5,000 tasks/mo
- 30-day retention
- Custom domain
- SSO
- SLA
Pro
Growing teams
- 25 agents
- 50,000 tasks/mo
- 90-day retention
- Priority
- Custom domain
- SSO
- SLA
Enterprise
At scale
- Unlimited
- Unlimited
- 1-year retention
- Dedicated
- Custom domain
- SSO
- SLA
Frequently asked questions
Ready to put your agents to work?
Create your workspace in two minutes. Free to start, no credit card required.