777genius/agent-teams-ai

You're the boss, agents are your team. They handle tasks on their own, message each other, and review each other's work. You just watch the kanban board and give high-level commands. Codex/Claude/OpenCode/Cursor/Grok/GitHub/Kiro/Z.AI/Xiaomi/MiniMax/Kimi(300+ models, 200+ LLM providers, free models no auth) Build your AI company with multiple teams

Agent Teams AI – Overview

What it is – A free, cross‑platform desktop application that lets you build teams of AI agents and watch them work on software projects. You create a team with a single prompt that defines each agent’s role, then the app runs the agents in parallel, lets them talk to each other, create and complete tasks on a Kanban board, and even review code changes.

Key capabilities

  • Multi‑model support – Works out‑of‑the‑box with a free, no‑auth model and can auto‑detect local runtimes such as Claude Code, Codex, OpenCode. You can also connect dozens of commercial providers (Cursor, SuperGrok, GitHub Copilot, Z.AI, MiniMax, Kiro, etc.).
  • Team orchestration – Agents are organised into teams; each team has its own Kanban board, chat, and runtime. Agents can create sub‑tasks, delegate, and communicate across teams.
  • Human‑in‑the‑loop controls – Per‑action approvals, role‑based permissions, and notification settings let you intervene when an agent wants to run a tool or make a change.
  • Rich UI – Integrated terminal (PTY) per team, built‑in code editor with Git support, task‑level logs, attachment handling, and a visual overview of organizations, squads, and agents.
  • Analytics & budgeting – Token usage, cost estimates, and budget alerts are displayed per task, model, and team, helping you keep expenses under control.
  • Extensibility – Includes an mcp-server component for plugging in external tools or custom agent plugins.

Installation – No build step required. Pre‑compiled binaries are provided for macOS (Apple Silicon & Intel), Windows (.exe), and Linux (AppImage, .deb, .rpm, .pacman). The installer can auto‑detect local runtimes; otherwise you add API keys for cloud providers via the UI.

Typical workflow

  1. Create a team – Provide a high‑level prompt describing the project and the roles you need (e.g., “frontend developer”, “tester”).
  2. Watch the Kanban – Tasks appear, change status, and agents log their actions in real time.
  3. Interact when needed – DM an agent, approve a tool call, or edit code directly in the built‑in editor.
  4. Review & merge – Agents perform code reviews; you accept, reject, or comment on the changes.
  5. Analyse – After a run, inspect token usage, CPU/RAM history, and session logs to understand cost and performance.

Who it’s for – Developers who want to experiment with autonomous AI‑driven development without writing a lot of glue code, teams looking for a visual way to coordinate multiple LLMs, or anyone curious about “AI‑as‑a‑team” workflows.

Links

Related

  • Project
    BloopAI/vibe-kanban

    A management layer for AI coding agents that combines kanban planning with dedicated workspaces for agent execution, review, and app previewing.

  • Project
    OrchestratorInc/agent-orchestrator

    A desktop workspace that coordinates multiple AI coding agents by providing a unified Kanban interface, isolated execution environments, and a project-level orchestrator for planning and delegation.

  • Project
    agent-tower/core

    A command center for managing multiple AI agents, featuring a Kanban dashboard, automatic Git Worktree isolation for conflict-free coding, and multi-agent collaboration.

  • Project
    cline/kanban

    A specialized IDE replacement for orchestrating multiple AI agents in parallel, using ephemeral worktrees and linked task cards to automate complex coding workflows.

  • Project
    kdlbs/kandev

    Kandev is an open-source control plane for orchestrating multiple AI coding agents in parallel, with kanban task management, multi-agent workflows, and an integrated review workspace. It solves the problem of scaling agent-based development while keeping humans in control of what gets shipped.