AI Blocks › Test an idea before you build it
Idea validation analyst
Give it your idea. It picks the riskiest guess, plans the test, reads the result and says keep, change or stop.
Idea validation analyst 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 people test an idea once, get a mixed result and read it the way they hoped. Some keep testing for months. Others stop after one bad answer from the wrong people. This agent runs the tests one round at a time and keeps a log. It looks for a fact that could kill the idea first. Then it picks the riskiest guess that is still untested and plans the cheapest test for it. The pass line is written down before the test runs, so a result cannot be bent to fit. When you bring the result back, it counts what people did and gave up, not what they said. A fail with an obvious fix gets one changed test. Two fails on the same guess, or four tests with no pass, get a keep, change or stop call. You decide.
You give it
- Your idea, who it is for, and any price or offer
- Results of the last test, such as notes from customer talks, sign-ups, pre-orders or replies
- What you can spend and how many weeks you will give each test (optional)
- Your notes (optional)
- Your results (optional)
- Your question (optional)
You get
- The riskiest guess and why it is first
- The next test, who to try it on, the pass line, the cost and the time
- Your last result judged against the line you set before it ran
- A keep going, change one thing or stop call, with the evidence
- A log of every guess, its test and its result
Use it when
- You have an idea and want to know if it is worth building before you spend on it.
- A test just finished and you are not sure what the result means.
- You have run a few tests and cannot tell whether to keep going.
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 give it your idea.
- Run the test it plans. Do not move the pass line while it runs.
- When the test finishes, run the agent again and give it the result. It reads the log and picks the next step.
- Read the call. If it says stop or change, check the evidence it lists before you decide.
- It stops planning after four tests with no pass. After that, you decide what is next.