esmirnov.aiUse this for science claims you want to check before spending money or changing a habit. The skill keeps a research brief ready for the next question, so you do not have to request every source check again.
Why use it?
It asks the assistant to read study methods, look for contrary findings and explain uncertainty. You get a short answer, evidence notes showing what was read, and a separate report of Reddit and forum experiences.
ChatGPT and Claude can do this with suitable instructions. The skill saves those instructions, including an editable source policy and a consistent report format. The example below shows why checking the methods matters.
Installation
Download the repository and keep the whole science-research folder, including references/:
git clone --depth 1 https://github.com/eduard-s/skills.git ai-skills
For a fresh personal installation, choose your app:
Claude Code
mkdir -p ~/.claude/skills
cp -R ai-skills/skills/science-research ~/.claude/skills/
Codex
mkdir -p ~/.agents/skills
cp -R ai-skills/skills/science-research ~/.agents/skills/
Claude web or desktop: ZIP the science-research folder. Upload it under Customize > Skills and enable it. Claude requires Code execution and file creation for Skills. See the upload guide.
ChatGPT desktop: Open Skills and use the skill-creation flow with the folder's Markdown files. You can ask @skill-creator to create the skill while preserving its references. See the official guide. This repository supplies a standalone skill, without a web/mobile plugin package.
Native Claude and ChatGPT imports have not yet been tested in this repository.
Use it
Invoke /science-research in Claude Code, $science-research in Codex, or select the installed skill with @ in ChatGPT. Then ask your question.
Browsing is needed for a current evidence search. The default output is a scientific brief, detailed notes and a separate experience report. Add science only to skip forums.
Try it
Use science-research.
What does research say about blue-light glasses for screen-related eye strain?
Sample answer
Blue-light glasses have not shown a reliable extra benefit over clear lenses for short-term screen-related eye strain. The controlled evidence is limited, so this does not settle whether a particular person might find a pair more comfortable.
The Cochrane review rated the evidence for subjective fatigue low certainty. A newer survey reported improvement in 85% of respondents, but surveyed existing users without a clear-lens control group. That percentage cannot tell us how much improvement the filter caused.
These findings give little reason to pay extra specifically for a promised eye-strain benefit. Longer-term effects remain uncertain.
The useful check here is the control group: an encouraging percentage means less once you see how the study was done.
Read the full walkthrough and comparison · Open the skill on GitHub