Form agent teams. Assign roles. Orchestrate multi-agent workflows with dependency tracking — all through the Model Context Protocol.
Subscribe — $19/month14-day free trial · Cancel anytime · No credit card lock-in
AgentTeam lets you define teams with specific roles, assign AI agents to those roles, and orchestrate task workflows with dependency tracking, status management, and progress reporting. All data is stored locally in SQLite — no cloud dependency.
MCP Native Claude Desktop VS Code Cursor SQLite
Create teams with custom names, goals, and role structures. Define who does what before any agent starts.
Assign specific agents to roles with capability requirements. Each role maps to an agent with verified skills.
Define tasks with dependency chains. Agents pick up work when their dependencies are satisfied — never out of order.
Track every task across every team. Statuses: pending, in-progress, blocked, completed. Full history.
All data in ~/.agentteam/ SQLite. No cloud. No latency. Your team data stays on your machine.
Integrates with Claude Desktop, Cursor, VS Code, and any MCP-compatible host via standard STDIO transport.
# Clone and install
cd agent-team-mcp
pip install -r requirements.txt
# Run as MCP server
python server.py
{
"mcpServers": {
"agent-team": {
"command": "python",
"args": ["/absolute/path/to/agent-team-mcp/server.py"]
}
}
}
| Tool | Description |
|---|---|
team_create | Create team with name, goal, roles |
team_assign_role | Assign agent to role with requirements |
team_list | List all teams and their status |
team_status | Get detailed status of a team |
task_create | Create task with dependencies |
task_assign | Assign task to agent/role |
task_update_status | Update task progress |
task_list | List tasks with filtering |
AGENTTEAM PRO
Unlimited teams · Unlimited agents · Priority support · 14-day trial
Subscribe NowAgentTeam MCP is part of the AgentPay Labs ecosystem — 28+ developer tools and MCP servers built for the agent economy. Every tool ships with a Stripe payment link, SEO-optimized landing page, and cross-linked ecosystem footer for organic discovery.