Open-source AI agent harness built with humans in mind.
License
GPL-3.0
Stars
9,285
Homepage
https://tinyhumans.ai/openhumanFeatures
- Open-source agent execution harness
- Human participation and supervision focus
- Built for observable workflows
- Tool-connected task execution
- Useful for agent prototype validation
- Fit for controlled automation scenarios
Use Cases
- Agent prototype development
- Human-agent collaborative automation
- Tool-calling workflow validation
- Internal task execution framework setup
- Supervised agent experiments
- Open-source AI engineering learning and extension
FAQ
OpenHuman is an open-source AI agent harness focused on keeping humans in the loop of agent execution. It helps developers build observable, interruptible, and collaborative automation workflows, making it useful for agent prototypes, tool-connected tasks, reusable execution flows, and supervised real-world agent operations. Repository: https://github.com/tinyhumansai/openhuman.
Common real-world fits include: Agent prototype development, Human-agent collaborative automation, Tool-calling workflow validation.
