Quickstart for building a grounded, governed agent workflow with traces, evals, and safe defaults.

Getting started

Overview

This guide walks through the minimum setup to run a production-oriented workflow: a read-only agent grounded on knowledge, with trace capture and an evaluation baseline.

What this documentation covers

  • Environment setup and authentication basics.
  • Create a first read-only workflow with citations.
  • Add a tool with a strict schema (read-only).
  • Instrument traces and cost metrics.
  • Build a small golden set for regression testing.

Who this is for

  • Engineers new to the platform who want a working baseline quickly.
  • Teams planning an incremental rollout from read-only to approvals.

Operational expectations

  • Examples favor deterministic behavior and explicit contracts (schemas, IDs, correlation).
  • Security is treated as a default: least privilege, audit logs, and safe fallbacks.
  • Changes are versioned; breaking changes should be announced via /docs/changelog.