Glow Documentation

Documentation for the Glow conversational AI training platform — an academic, self-hostable system for designing and running persona-driven simulations with full grading + analytics.

Getting started

• Start — install + first-boot of a Glow instance
• How It Works — the simulation, persona, scenario, attempt loop
• Tutorial — design your first simulation end-to-end
• Patterns — proven patterns + anti-patterns
• Annotated Example — a real seed walked through line-by-line
• Authentication — sessions and bearer tokens

Reference

• API Reference — every HTTP endpoint, auto-generated from OpenAPI
• CLI Reference — every glow command, auto-generated from the CLI spec

Sidebar layout

The sidebar is organized by API/CLI taxonomy — each group corresponds to a parent artifact or a logical cluster of artifacts in the Glow API. The client product UI uses its own grouping (Learner / Analytics / etc.) optimized for product navigation; the two intentionally differ.