hive
Multi-Agent Harness for Production AI
Hive is a multi-agent runtime harness designed for production AI workloads, orchestrating colonies of specialized agents that work together on business processes. The system features a Queen agent that leads and spawns worker agent clones on demand, coordinating through a shared tracker ledger without requiring manual workflow graphs. Built with crash-safe state management, cost enforcement, observability, and human-in-the-loop oversight, Hive supports 100+ LLM providers including OpenAI, Anthropic, Google Gemini, and local models. The framework enables agents to improve over time through reflexion, scoped memory, and learned skills while providing browser automation and general compute capabilities.
Key Features
Use Cases
- 01Automating multi-step business processes that require parallel execution and state persistence
- 02Running long-running agent workflows with automatic crash recovery and resume capabilities
- 03Deploying production AI agents with strict budget controls and audit trail requirements
- 04Building adaptive agents that learn and improve task execution over time through reflexion
- 05Orchestrating browser automation and web-based workflows across multiple parallel sessions
- 06Implementing human-in-the-loop AI workflows with escalation policies and approval gates
Related Agents
View morehermes-agent
The agent that grows with you
agency-agents
A complete AI agency at your fingertips - From frontend wizards to Reddit community ninjas, from whimsy injectors to reality checkers. Each agent is a specialized expert with personality, processes, and proven deliverables.
openinterpreter
A coding agent for open models like Kimi K3
cline
Autonomous coding agent as an SDK, IDE extension, or CLI assistant.
hive — FAQ
What is Hive and what does it do?+
Hive is a multi-agent runtime framework that orchestrates colonies of AI agents to execute production business processes. It features a Queen agent that leads and spawns worker clones dynamically, coordinating through a shared tracker ledger with built-in state management, cost controls, and human oversight.
How do I install and set up Hive?+
Clone the Hive repository and run the quickstart script (./quickstart.sh on macOS/Linux or .\quickstart.ps1 on Windows PowerShell). The setup configures the framework, installs MCP tools, sets up credential storage, and prompts you to configure your LLM provider. Note that Hive requires Python 3.11+ and uses a uv workspace layout, not pip install.
Which LLM providers does Hive support?+
Hive supports 100+ LLM providers through LiteLLM integration, including OpenAI, Anthropic Claude, Google Gemini, DeepSeek, Mistral, Groq, OpenRouter, and Hive LLM. It also supports local models via Ollama using the format ollama/model-name.
Do I need API keys to use Hive?+
Yes, you need API keys from your chosen LLM provider(s) such as OpenAI, Anthropic, or Google. The quickstart script guides you through configuring your default model and storing credentials securely in the encrypted credential store.
Is Hive free and open-source?+
Yes, Hive is fully open-source and licensed under Apache License 2.0. The framework is free to use, though you'll need to pay for LLM API usage from your chosen provider.
Does Hive support human-in-the-loop workflows?+
Yes, Hive includes Sentinel, an out-of-band human-in-the-loop system. Agents can escalate to humans via Slack or Telegram, parking their state until receiving a response, then resuming exactly where they left off with configurable timeouts and escalation policies.
How do I install hive?+
Open the source repository on GitHub and follow its README. hive is a agent — MCP Agents Market links you directly to the official repo.
Is hive free?+
hive is an open-source project hosted on GitHub. Check the repository for its license and any usage requirements.