AI Blocks › Test an idea before you build it
Product experiment designer
Give it an idea to test. Get a clear hypothesis, sample size, test plan and how to read the result.
Product experiment designer is a free AI skill for the job "Test an idea before you build it". Give the instructions to an AI assistant that supports the required files and tools, then review its output.
Free · No account · MIT licence · Made by Alireza Rezvani
Made by Alireza Rezvani. See the original on GitHub; we copied it as it is, with its licence.
What it does
A test plan you can run and trust: a clear hypothesis, the right sample size, and rules for when to stop and how to read the result. It keeps you from calling a winner too early. Claude writes the idea in If, Then, Because form. It picks one decision metric, adds guardrail metrics to catch harm, and treats other numbers as diagnostics only. A Python script works out sample size from your baseline rate, the smallest change worth detecting, confidence and power. Ideas are ranked with ICE scoring: impact, confidence and ease. It then sets stopping rules before launch and reads the outcome.
You give it
- The change you want to test
- Your current conversion rate or average
- The smallest improvement that would matter
- Test results to interpret (optional)
You get
- A written hypothesis
- Primary and guardrail metrics
- A sample size and test duration
- ICE ranking of test ideas
- A plain reading of the result
Use it when
- You have several ideas to test and need to pick the order.
- You want to know how many users a test needs.
- A test finished and you are unsure if the result is real.
- You need a simple hypothesis the whole team can agree on.
How to use it
- Add it to your AI once: download it and add the folder to Claude, ChatGPT or Codex as a skill.
- Ask for the job in your own words. Your AI opens the skill when it fits.
- It works best when your AI can run code.