Concepts

Deep dive into Goa-AI’s core concepts: DSL, runtime, toolsets, and MCP integration.

This section covers the core concepts you need to understand to effectively use Goa-AI:

  1. DSL Reference - Learn all available DSL functions for defining agents, toolsets, and policies
  2. Runtime - Understand how the runtime orchestrates agents, enforces policies, and manages state
  3. Toolsets - Learn about toolset types, execution models, and composition patterns
  4. MCP Integration - Integrate external MCP servers into your agents
  5. Transcripts - Learn how transcripts and message parts simplify planner, UI, and provider state
  6. Tool Validation & Retry Hints - See how Goa-AI turns validation failures into structured retries the planner can repair
  7. Tool Catalogs & Schemas - Discover tools and JSON Schemas exported by agents
  8. Run Trees & Streaming - Understand per-run streams, AgentRunStarted links, and stream profiles
  9. Policies & Labels - Configure caps, time budgets, and labels via RunPolicy and runtime overrides
  10. LLM Integration - Integrate LLM providers like AWS Bedrock and OpenAI through the provider-agnostic model client interface

In This Section

DSL Reference

Complete reference for Goa-AI’s DSL functions.

Runtime Concepts

Understand how the Goa-AI runtime orchestrates agents, enforces policies, and manages state.

Toolsets

Learn about toolset types, execution models, and composition patterns.

MCP Integration

Integrate external MCP servers into your agents.

Transcripts & Message Parts

Understand Goa-AI’s transcript model and how it simplifies planner and UI state.

Tool Validation & Retry Hints

Learn how Goa-AI turns validation failures into structured ToolError and RetryHint values that LLMs can use to repair tool calls.

Tool Catalogs & Schemas

Discover tools, schemas, and specs exported by Goa-AI agents, and learn how UIs and planners can consume them.

Run Trees & Streaming Topology

Understand how Goa-AI models agent runs as a tree and how streaming projects that topology to different audiences.

Policies, Caps & Labels

Learn how Goa-AI enforces caps, time budgets, and policies around agent runs using design-time RunPolicy and runtime overrides.

LLM Integration

How Goa-AI integrates with LLM providers through a provider-agnostic interface and adapter modules.