AI-written blog paragraph: 'A 2024 Stanford study found 73% of shoppers abandon carts over shipping costs.'
You get
Claim: '2024 Stanford study, 73%' — unsupported, high risk (specific number and named institution).
Verify: find the original study; cart-abandonment figures usually come from Baymard Institute surveys, not Stanford.
Safer: cite the study you actually find, with its year and sample.
About this skill
About AI draft fact-checker
AI drafts sound confident even when names, numbers, quotes or citations are wrong, and errors get published. This skill lists every checkable claim in your draft — numbers, dates, names, quotes, citations and legal or medical statements — and marks each as supported, unsupported or contradicted by the sources you paste. It rates the risk if a claim is wrong, explains how to verify each high-risk one and suggests safer wording. It never verifies a claim from memory, and treats any citation it can't see as unverified. To use it, paste the skill into your assistant and send a draft written with (or by) AI, and the sources it was based on if you have them.
AI draft fact-checker — example input and the first lines it returns. Works in ChatGPT, Claude and Gemini.
2 Paste it into a new chatChatGPT, Claude, Gemini or Copilot.
3 Send your inputSend a draft written with (or by) AI.
The skill — paste this into your AI assistant
# Skill: AI draft fact-checker
You are running the "AI draft fact-checker" skill for me. I work as: Writer or editor publishing AI-assisted work. What I need from you: every checkable claim in my draft listed, rated by risk, with what to verify and where.
## When I send you a draft written with (or by) AI, and the sources it was based on if I have them
1. Extract every factual claim: numbers, dates, names, quotes, citations, legal or medical statements.
2. Mark each as supported by the sources given, unsupported or contradicted.
3. Rate risk if wrong (high, medium, low).
4. Say exactly how to verify each high-risk claim and suggest safer wording.
## Reply in this format
- Claims table: claim, status, risk, how to verify
- High-risk claims first
- Safer rewrites
- What you can publish as-is
## Rules
- Never mark a claim verified from memory; only from sources the user pasted.
- Treat any citation you can't see as unverified.
- Use only the facts I give you; mark anything you had to assume with [ASSUMPTION].
- If something essential is missing, ask me one short question, then continue.
- No preamble or filler: start with the result.
Confirm you're ready in one line, then wait for me to send a draft written with (or by) AI, and the sources it was based on if I have them.
Then sendA draft written with (or by) AI, and the sources it was based on if you have them. For example: AI-written blog paragraph: 'A 2024 Stanford study found 73% of shoppers abandon carts over shipping costs.'
ChatGPT
Fastest: paste the skill as your first message in a new chat, then send your input.
To keep it: open a Project, choose Instructions (or Settings → Personalization → Custom instructions for every chat) and paste it there.
To share it with a team: create a custom GPT and paste the skill into its Instructions.
In the Codex app: copy the skill's folder into ~/.codex/skills/ (or Customize → Skills → Add → Create skill).
Claude
Fastest: paste the skill as your first message in a new chat, then send your input.
To keep it: create a Project, open Project instructions and paste the skill.
As a Claude Skill: download the skill's ZIP, then in Claude open Settings → Skills → Add → Upload skill and choose the file.
Gemini
Fastest: paste the skill as your first message in a new chat, then send your input.
To keep it: open Gems → New Gem, paste the skill into Instructions and save.
Microsoft Copilot
Fastest: paste the skill as your first message, then send your input.
In Microsoft 365 Copilot, save it as a prompt or paste it into an agent's instructions to reuse it.
Any other assistant
Paste the skill as the system prompt or first message in any chat assistant (Perplexity, Meta AI, DeepSeek, Mistral, local models).
Coding agents that read the Agent Skills format (SKILL.md) can load the downloaded file from their skills folder.
Install it permanently
Upload it to Claude, add it to ChatGPT's Codex app, clone it into Claude Code, or download the SKILL.md for any agent. Full guide with screenshots →
Download the ZIP and upload it in Claude. Claude then uses the skill on its own whenever a request matches.
Claude Code: save it as ~/.claude/skills/ai-draft-fact-checker/SKILL.md.
Claude apps: use the Claude app tab — it has a ready-made ZIP.
Other agents: place the folder wherever your tool loads Agent Skills from.
Repositoryscienceofapps.com/skills/ai-skills.git · all 200 skills, one folder each, plain git over HTTPS
What it does
The steps it follows
Extract every factual claim: numbers, dates, names, quotes, citations, legal or medical statements.
Mark each as supported by the sources given, unsupported or contradicted.
Rate risk if wrong (high, medium, low).
Say exactly how to verify each high-risk claim and suggest safer wording.
What you get back
The reply format
Claims table: claim, status, risk, how to verify
High-risk claims first
Safer rewrites
What you can publish as-is
Rules it never breaks
Never mark a claim verified from memory; only from sources the user pasted.
Treat any citation you can't see as unverified.
Uses only the facts you give it and marks assumptions.
FAQs
Questions about AI draft fact-checker
What does the AI draft fact-checker skill do?
AI drafts sound confident even when names, numbers, quotes or citations are wrong, and errors get published. The skill fixes that: every checkable claim in your draft listed, rated by risk, with what to verify and where.
How do I install AI draft fact-checker in ChatGPT, Claude or Gemini?
Copy the skill and paste it as the first message of a new chat, then send your input. To keep it, paste it into a ChatGPT Project or custom GPT, a Claude Project or a Gemini Gem, or upload the SKILL.md file as a Claude Skill.
What do I send AI draft fact-checker after installing it?
Send a draft written with (or by) AI, and the sources it was based on if you have them. If something essential is missing, the skill asks you one short question and then continues.
Is the AI draft fact-checker skill free?
Yes. It's free to copy, use and adapt, with no sign-up, and works in the free versions of ChatGPT, Claude, Gemini and Copilot.
60 research-backed principles in 6 categories — what each one means, why it works on the human brain and how to use it on your own screens. Written by Sushant Singh, founder of Science of Apps, who has scaled consumer apps to lacs of users and spent 10+ years leading product across startups and digital businesses in the US, EU & India.
Free. No card, no paywall — just your email.
Free PDF · 90 pages · 2 editions
100+ Subscription Paywalls, decoded
110 real paywalls from top apps, taken apart, rated out of 5 and decoded with the Laws of UX — plus the frameworks, copy and checklists to ship your own. By Sushant Singh, founder of Science of Apps.
110 paywalls rated
34 Laws of UX
12 frameworks
Which paywall scored 4.8 out of 5, and which scored 2.6? It’s all inside.