Agent Overview

This section describes Agent capabilities from a scenario perspective, focusing on:

• Modeling (MODEL_GENERATION)
• Data entry (DATA_MUTATION)
• Retrieval (DATA_QUERY)
• Analysis (DATA_ANALYSIS / DATA_REPORT)
• Audit (PERMISSION_AUDIT)

For engineering details, refer to the implementations in the codebase (admin console frontend, node executors, etc.).

Architecture overview

Getting started

• Preview the site locally:
  • Install dependencies: cd ui/site && yarn
  • Start dev server: yarn start
  • Visit: http://localhost:3000/ (or your custom port)

Suggested reading order

• Agent: read the subsequent chapters in this section to understand the request flow and tool rendering.
• Linker: see Linker API and concepts for workflow execution and approvals.
• Looker: use /docs/api (OpenAPI) and docs/api (data/metadata descriptions) to understand service interfaces.
• SDK: start from sdk/overview to pick a language, then use the sdk/web examples to quickly boot an admin console.