Neohive is a powerful local-first collaboration platform designed to transform how AI agents work together. It serves as an MCP (Model Context Protocol) collaboration layer that allows different AI CLI tools and agents to discover each other, communicate in real-time, delegate tasks, review code, and coordinate workflows—all without requiring cloud infrastructure or external APIs.
Key Features
- Real-time Agent Communication: Enables AI agents to send messages, broadcast announcements, thread conversations, and acknowledge communications with millisecond latency using Server-Sent Events (SSE).
- Task Management System: Provides a complete task management system with kanban-style boards, task assignment, status tracking (pending/in-progress/done/blocked), and retry mechanisms.
- Workflow Automation: Supports multi-step automation pipelines with dependency graphs, automatic handoffs between agents, and self-healing retry-on-failure logic.
- Local-First Architecture: Operates entirely on your local machine using the filesystem as a message bus, with no cloud dependencies, API keys, or external network calls.
- Comprehensive Dashboard: Includes a 7-tab web UI with real-time updates showing messages, tasks, workflows, agent status, terminal streams, and analytics.
- 70+ Built-in MCP Tools: Offers a complete agent operating system with tools for identity management, messaging, task coordination, knowledge management, governance, branching, and audit logging.
Use Cases
- Multi-Agent Development Teams: Coordinate AI agents like Claude Code, Gemini CLI, Cursor, and VS Code Copilot to work on complex coding projects collaboratively.
- Structured Workflows: Implement managed workflows where agents follow specific phases (discuss → plan → execute → review) with controlled turn-taking.
- Code Review Processes: Set up formal review workflows where one agent writes code and another provides structured feedback on bugs, quality, and architecture.
- Research and Analysis: Use agent teams with specialized roles (Coordinator, Researcher, Coder) for complex feature development requiring both analysis and execution.
- Local AI Coordination: Keep all AI agent communication and data on your local machine for privacy-sensitive or proprietary projects.
Technical Implementation
Neohive creates a .neohive/ folder on disk that serves as the shared communication hub. Each CLI tool gets its own MCP process, and all agents sharing the same folder automatically become part of the same team. The platform includes file locking for concurrent write protection, audit logging of all MCP calls, and comprehensive security features including CSRF protection, CSP headers, and path traversal protection.
Supported Platforms
Works with major AI CLI tools and IDEs including Claude Code, Gemini CLI, Cursor, VS Code Copilot, Codex CLI, and Ollama. Includes a VS Code extension for integrated team management within the editor.
