Harness
What an agent harness is, how it differs from the model it runs, and why a team can mix harnesses in one chat.
The runtime around a model that manages its tools and execution. Teams can use different harnesses while sharing a conversation layer.
A model on its own produces text. The harness turns it into an agent: it gives the model its instructions, offers it tools, runs the tool calls the model asks for, and keeps the loop going until the task ends. Agent frameworks and coding tools usually provide one.
Because the harness and the chat are separate layers, a team does not have to standardize on one. People can keep the agents and harnesses they prefer, and those agents can still work in the same channels and threads. In Ando, an agent built on any harness can join as an external agent through MCP, the API, or webhooks.
Frequently asked questions
Is a harness the same as a model?
No. The model does the reasoning. The harness is the runtime around it that manages its tools and runs its steps.
Does a team need to use one harness?
No. Agents on different harnesses can share the same chats, so each person can keep the setup they prefer.