FAQ
What is Agiwo?
Section titled “What is Agiwo?”Agiwo is a Python runtime harness for orchestrated, tool-using, traceable, and increasingly self-improving agent systems. The SDK is the execution core; the Console is an optional self-hosted control plane.
What does “runtime harness” mean?
Section titled “What does “runtime harness” mean?”It means Agiwo treats execution control, orchestration, persistence, traces, memory, skills, and operator visibility as first-class runtime concerns rather than incidental details behind prompts.
How is Agiwo different from LangGraph?
Section titled “How is Agiwo different from LangGraph?”Agiwo is built around a canonical agent runtime plus scheduler orchestration. LangGraph is built around graph-native workflow authoring. If you want explicit runtime boundaries and a control-plane story, Agiwo is usually the better fit.
Does Agiwo support multi-agent systems?
Section titled “Does Agiwo support multi-agent systems?”Yes. Agiwo supports both Agent.as_tool() composition and scheduler-managed child-agent workflows.
Is the Console required?
Section titled “Is the Console required?”No. The SDK can be used on its own. The Console is a separate control plane.
Does Agiwo include persistence and traces?
Section titled “Does Agiwo include persistence and traces?”Yes. Run and step storage plus trace storage are first-class parts of the runtime model.
What does self-improving mean here?
Section titled “What does self-improving mean here?”It means Agiwo is designed for feedback loops such as memory retrieval, skills, context rollback, tool-result retrospect, and operator-visible traces. It does not imply unconstrained autonomous code mutation.