How it works
- Connect GrowthBook's MCP server to your AI assistant or coding agent following the GrowthBook MCP documentation, using an API key with read access.
- Install the ExpLint MCP at https://explint.com/mcp to enable your agent to audit tests.
- Run the prompt below to audit these plans before the work starts.
How the check runs
- It reads the draft's name, hypothesis, description, variations, metrics, and traffic split, then builds one text that keeps your team's headings so the draft and the lint report line up field by field.
- It sends that text to ExpLint with
explint__lint, which returns errors, warnings, and passes, each finding naming the rule behind it. - You edit the draft in GrowthBook and ask for a recheck, and the assistant confirms which findings cleared and which remain.
Copy and paste this prompt
Change the words in square brackets to name your own project, board, or test.
Prompt for your assistant
Use GrowthBook MCP tools to read draft experiment [experiment name]. Build one text from its hypothesis, description, variations, primary and guardrail metrics, traffic split, and run length. Lint it with explint.com against project [your ExpLint project name]. Show each error and warning with its GrowthBook field. Do not change GrowthBook yet. Suggest edits here, and on my update, read and lint the draft again.
What ExpLint looks for
ExpLint ships over 24 rules, and a project switches each one on or off or rewrites it in plain English. These are the rules the plans this guide reads break most often. An error means the test cannot answer its question as written, and a warning means the result will be hard to judge when it comes in.
| Rule | Level | In plain words |
|---|---|---|
| hypothesis/names-metric-and-direction | error | The hypothesis field has to name the metric and the direction it expects, since GrowthBook records metrics separately and the prose is where the expectation lives. |
| variants/control-is-described | error | GrowthBook names a baseline variation, and the draft still has to say what that baseline shows people today for a new reader to picture it. |
| power/has-a-sample-size-calculation | error | A draft that sets a 50/50 split with no headcount leaves the team unable to say when the test has seen enough traffic. |
| duration/covers-full-business-cycles | error | The run setting should cover full weeks, because a Monday to Friday window reads the same traffic every day of its life. |
