Introducing the Gainsight AI Cookbook for Admin CLI

Most admins inherit a tenant they didn't build. Hundreds of rules, dozens of CTAs, program logic nobody documented, and a Slack thread from 2023 that half-explains why a field exists. Answering "what does this tenant actually do?" usually means clicking through config screens for a week.

We built something better, and today we're making it public.

The Gainsight AI Cookbook

https://github.com/Gainsight-Central/gainsight-ai-cookbook

First kit: CS Admin CLI Starter Kit

Point an AI coding agent at your CS tenant through the Admin CLI and ask questions in plain English instead of reading config pages. Treat this as a starting point, not a finished product. Drop it in your working folder and it gives your AI agent the scaffolding it needs to answer Gainsight questions properly: what the objects are, how they relate, which commands to reach for. From there you make it your own by adding your team's conventions, your naming standards, your recurring questions.

What it gives you

• A set of read-only prompts, so nothing in the kit can change your tenant
• An optional Claude skill that teaches the agent how Gainsight CS objects fit together
• Setup instructions that get you running in a few minutes

A use case: onboarding onto an unfamiliar tenant

You've just taken over a Gainsight instance. Instead of a week of discovery, you run the starter kit and ask:

• “What high level risk plays do we have? How is risk detected and acted?”
• "Which rules write to this field, and what how frequently they run in?"
• "Which Journey Orchestrator programs are active but haven't sent an email this quarter?"

You get a written map of your own configuration, produced by something that can read all of it at once. Same pattern works for cleanup projects, audit prep, and pre-migration inventory.