Reference
Auto-generated API reference for the flowai_harness public surface. Each page below renders the module's exported classes, functions, and type aliases directly from their...
Auto-generated API reference for the flowai_harness public surface. Each
page below renders the module's exported classes, functions, and type
aliases directly from their docstrings.
For task-oriented guides see Approvals, Data Environment, Studio, Expose Tools Over MCP, Testing, and Streaming Events.
Besides dev / serve (see Studio), the CLI also provides
flowai-harness docs export to generate the docs content artifact and
flowai-harness mcp python MODULE:OBJECT to serve a Python runtime object
over MCP (see MCP).
- Runtime —
Runtime,create_runtime,define_runtime,RuntimeSpec, and the testing / data-environment configs. - Studio —
define_app, workspace bindings, and the local Studio command surface. - Evals — eval configs, test cases, structured action ground truth, scorer presets, artifacts, and event envelopes.
- MCP — stdio and Streamable HTTP helpers for exposing runtime tools to MCP clients.
- Agents —
define_coordinator,define_planner,define_executor,define_specialist, andAgentSpec. - Plans —
define_plan,PlanSpec,PlanDisplayAlias. - References —
define_reference,ReferenceSpec. - Tools —
define_tool,ToolSpec. - Prompts —
layered_prompt,LayeredPrompt. - Tenant —
define_tenant,TenantIdentity. - Unions —
TaggedUnion. - Glimpse —
glimpse.
Streaming Events
runtime.query(prompt, thread_id=...) and runtime.run_specialist(specialist, prompt, thread_id=...) both return an async iterator over wire-shape event dicts. Each event is a...
Runtime
Native-backed runtime handle, spec constructors, approval and storage policy values, and testing / data-environment configuration.