AI Blocks › Research a topic properly
Research hypothesis builder
Describe what you observed. Get testable hypotheses, rival explanations, predictions and a plan to test them.
Research hypothesis builder is a free AI skill for the job "Research a topic properly". 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 K-Dense · Thinking: high
Made by K-Dense. See the original on GitHub; we copied it as it is, with its licence.
What it does
Testable explanations for something you observed, with rival explanations, predictions that tell them apart and a plan to test them. It keeps you from falling for the first story that fits. Starting with your observation, it writes a clear question, then several candidate hypotheses. For each it states what would have to be true, what rival explanation could explain the same data, and what result would tell them apart. It keeps association and cause separate and never claims a hypothesis is proven or new. It includes small local Python tools that handle your files without a network. It stops at ethics, safety and privacy gates and does not rank ideas for you.
You give it
- Your observation or data
- Field and context
- Existing literature notes (optional)
- Constraints on what you can measure (optional)
You get
- Clear research question
- Candidate hypotheses
- Rival explanations
- Predictions that separate them
- Measurement and testing plan
Use it when
- Your data shows a pattern and you want to explain it.
- You want rival explanations before you design an experiment.
- A result surprised you and you need to know what to test next.
- You are unsure if a link is cause or just association.
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.