How it works
- Connect the Convert MCP server at https://mcp.ai.app.convert.com/mcp to your AI assistant or coding agent, then sign in with Convert OAuth and choose the accounts and projects the assistant may read.
- 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 experience's description and objective, variations with the baseline marked, traffic distribution, audiences and locations, and the primary goal beside any other goals attached to the test, keeping those field names as headings.
- It sends that text to ExpLint with
explint__lint, which returns errors, warnings, and passes, each finding naming the rule behind it. - You review any suggested change in Convert before saving, since writes there are proposed for review and land only when you accept them, then ask for a fresh read.
Copy and paste this prompt
Change the words in square brackets to name your own project, board, or test.
Prompt for your assistant
Use the Convert MCP tools to read the experience [experience name]. Build one text from its description and objective, variations and traffic distribution, audiences, primary goal and other goals, and minimum and maximum running time. Lint it with explint.com against project [your ExpLint project name]. Show each error and warning with its Convert field. Do not launch or edit the experience yet. Suggest fixes here and re-check after I update Convert.
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 objective has to name the primary goal and the direction it expects, since Convert records the goal as a setting and the description carries the expectation. |
| variants/each-differs-from-control-stated | error | Each variation should state its own change, so a result that names one variation can be credited to something the reader can see. |
| randomization/states-the-split | error | The traffic distribution belongs in the plan text, because a share of visitors set in the interface is one the reader cannot judge against the runtime. |
| duration/covers-full-business-cycles | error | Convert plans runtime in days and visitor caps, and the window should cover full weeks for the weekday and weekend traffic to be read together. |
