AI Blocks › Test an idea before you build it
Experiment results review agent
Give it each test before it starts and when it ends. Get a readiness check, an honest verdict and the lesson saved.
Experiment results review agent is a free AI agent 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 · Made by AI Blocks · Thinking: high
What it does
Most tests fail before they start. Two things change at once, no rate is named, the site has a tenth of the traffic needed, and someone peeks on day three. Then a gap that is pure luck gets shipped and a real win gets missed because the readout was never done the same way twice. This agent sits at both ends. Before a test it defines the terms, checks the plan changes one thing with a written prediction, and scores readiness from your traffic. Under the bar, it sends the test back. When the test reaches its read date it reads the result with a confidence range and a split check, gets a second opinion on luck from a different reader, lists the side effects and says plainly what the test did not test. If the test was peeked at, stopped early or muddled, it writes the lesson as a…
You give it
- The test plan, with what changes, for whom, the rate to move, by how much and the read date
- Visits, conversions and revenue for the pages in the test, for the last few weeks
- The results per version when the test ends, by day if you have them
- What else happened during the test, with dates (advised)
- The test register, once it exists
- Your results (optional)
- Your question (optional)
- A buyer brief (optional)
You get
- The one change, the one rate and what would prove the idea wrong, written before the run
- A readiness score out of 100 with the sample size and weeks needed
- A verdict with both versions, the gap, the confidence range and a ship call
- A second opinion on luck and whether it agrees
- What the test did not test
- A lesson written as a rule when the test was run badly
Use it when
- Your team runs A/B or split tests and the readouts depend on who reads them.
- A test shipped last quarter and nobody can say now whether it worked.
- You want to know, before you build a test, whether you have the traffic to read it.
How to use it
- Start a chat with the agent. In Claude Code, add it as a skill or save it in .claude/agents, then ask it to check a test plan.
- Paste the plan and the traffic numbers. Read the readiness score. Fix the plan or drop the test if it says so.
- Start the test in your own tool. Tell the agent the start date.
- On the read date, export the results per version and give them to the agent. Read the verdict.
- Ship or revert yourself. The register keeps the team's hit rate over the year.