# Check a draft against its sources

A free AI skill from Neosaipien. Version 1.1, as of 16 October 2026.

## What it does

You give your AI assistant a draft (an email, a report, an update, written by AI or by you) and the documents it is based on. The skill checks every claim in the draft against those documents and gives you back:

- **A claim table.** Each claim is marked **Supported** (with the exact line from your source it rests on), **Partly supported**, **Contradicted**, or **Not in the sources**, with a suggested fix.
- **A "check by hand" list** of every name, number, date and quote in the draft, with the ones that differ put first.
- **A one-line verdict:** "Fix first", "Needs more sources", or "Ready to send, after the checks at the end".

It uses only the documents you give it. It never fills a gap from the web or from the AI's own memory: anything your documents do not say is reported as "Not in the sources". It suggests fixes but does not rewrite your draft unless you ask.

## Who it is for

Anyone who sends AI-assisted writing: managers, analysts, communications teams, salespeople and assistants.

## Why it matters now

- In the 2026 Work Trend Index (a survey of 20,000 knowledge workers who use AI at work, in 10 markets, published 5 May 2026), the two human skills rated most important were "quality control of AI output (50%)" and critical thinking (46%). [1]
- Anthropic's Economic Index report of 26 June 2026 found that "work conversations most often produce documents and reports (20%)". [2]

This skill is that quality check, written down so your assistant does it the same way every time. It goes with the "Check what AI writes" workshop in a box.

## What is in the download

| File | What it is |
|---|---|
| `check-against-sources.zip` | The skill, ready to upload to Claude |
| `skill/check-against-sources/` | The same skill as a folder: `SKILL.md` (the instructions) and `LICENSE.txt`. `SKILL.md` is the file for ChatGPT and Gemini |
| `portable/paste-in.txt` | The same instructions as plain text, for a plan without skills: paste it into a project's instructions, or at the start of a chat |

## Add it to your AI tool (as of Oct 2026)

All three tools now take skills in the same open format, a SKILL.md file. The format began at Anthropic and is now an open standard. So this is one skill, not a version per tool. Use the tool your company approves for work data.

If you cannot see a menu below, the usual reasons are a rollout still in progress, your plan, your country or an admin setting.

### Claude

**Plans:** Pro, Max, Team and Enterprise, says Anthropic's skills page (its pricing table also ticks Free).

1. In **Settings > Capabilities**, turn on **Code execution and file creation**. Skills need it.
2. Go to **Customize > Skills**.
3. Upload `check-against-sources.zip`. It holds the SKILL.md file.
4. To use it, describe the task in a chat, or type **/** and pick **check-against-sources**.

Anthropic does not review skills made by other people, so read a skill's files before you turn it on. This one is plain text: open `skill/check-against-sources/SKILL.md` to read it. A skill you add in Claude also follows you into Claude Code when you sign in with the same account.

### ChatGPT

**Plans:** skills are a beta on Free and Plus, and generally available on Enterprise and Edu (since 9 Jul 2026). ChatGPT keeps skills inside plugins, in the Plugin Directory.

1. Start a new chat and type **@skill-creator**.
2. Attach `skill/check-against-sources/SKILL.md` and write: "Make a skill from this file. Keep its name, description and instructions exactly as written."
3. Check that the skill it shows you matches the file.
4. To use it, describe the task, or type **@** and pick **check-against-sources**.

### Gemini

**Plans:** 18 or over, with a personal Google Account. Google says work and school accounts get skills "soon".

1. Open the left menu, click **Customize**, then **Skills**.
2. Make a new skill by uploading a SKILL.md: choose `skill/check-against-sources/SKILL.md`.
3. To use it, describe the task, or type **/** (Google says **@** is coming) and pick **check-against-sources**.

Gemini wants skill names in lowercase with hyphens, which this one has, and allows up to 100 active skills.

### No skills on your plan? Use a project

- **Claude or ChatGPT:** make a project for this job and paste the whole of `portable/paste-in.txt` into the project's instructions. Every chat in that project then follows it. In ChatGPT: select **New project** in the sidebar, name it, then add instructions from the **•••** menu under **Project settings**.
- **Gemini with a work account, or any single chat:** paste `portable/paste-in.txt` as your first message. It lasts for that one chat.

There is no custom GPT or Gem version: OpenAI plans to retire custom GPTs in favour of plugins, and Google is moving Gems to skills from Nov 2026 for personal accounts and in 2027 for work and school accounts.

## Which tool, mode and model to use

**My recommendation: Claude first, ChatGPT second, Gemini third.** (Samson, Neosaipien.) Checking a draft against your files is a plain chat job: no agent, no web search. What matters is a careful model and keeping your sources together.

| | Claude | ChatGPT | Gemini |
|---|---|---|---|
| Where to run it | Any ordinary Claude conversation | Chat | Chat |
| Model for an email or a short update | Sonnet 5.5 at its Default effort | GPT-6 at Medium | Flash |
| Model for a long report or many sources | Opus 5.5 | GPT-6 at High | Pro, thinking High |
| Keeping the sources together | A project holding the source files | A project holding the source files | A notebook in Gemini Notebook, built for answers drawn only from your own files |

- **Where to pick the model:** Claude, the model name next to the send button. ChatGPT, GPT-6 in the model menu, then the Thinking slider. Gemini, the current model at the bottom of the chat box.
- **Plans:** Opus 5.5 needs Claude Pro or above. GPT-6 needs ChatGPT Plus or above; Free and Go use GPT-6 Luna. Gemini's Flash and Pro need a Google AI plan; without one you get Flash-Lite.
- **Turn web search and deep research off for this job.** The point is to test the draft against your documents, not against the web. If you need new sources first, run deep research separately, then check its sources too.

## How to use it

1. Start a new chat.
2. Upload or paste your draft and the documents it is based on. Say which is which if the file names do not make it obvious.
3. Ask, for example: "Check this draft against these sources." or "Fact-check my email against the call notes and the product sheet."
4. Read the verdict first, then the rows marked Contradicted, Partly supported and Not in the sources.
5. If you want, ask "Now give me a corrected draft." It changes only what the table says, marks anything without a source as [needs a source], and lists every change.

## What you must still check

The skill ends every report with these four points, because they are yours to do:

1. That each "Supported" verdict really matches the quoted source line. Open the source and look.
2. That your sources are current and trustworthy. The skill cannot judge that.
3. Read the draft once yourself, for any claim the skill missed.
4. Anything marked "Not in the sources": check it somewhere reliable, or cut it.

## What it does not do

- It does not search the web, and it cannot tell whether a source is right or up to date.
- It does not check spelling, grammar, tone or legal wording, except where a name or figure differs from a source.
- It does not prove a draft is true. A "Supported" claim is only as good as the source behind it.
- It does not send, post or share anything, and it does not open links.
- It does not guarantee any result. You remain responsible for what you send.

## Your data

The skill is a set of written instructions. It runs no code, needs no internet connection and sends nothing anywhere. Your draft and your documents stay inside the assistant you use, under that assistant's own data terms. If your work has rules about which AI tools may see which documents, follow them.

## Compatibility, as of October 2026

| Tool | Who can add skills (vendor's own pages) | Which file |
|---|---|---|
| Claude | Pro, Max, Team and Enterprise (the pricing table also ticks Free), with code execution on | `check-against-sources.zip` |
| ChatGPT | A beta on Free and Plus; generally available on Enterprise and Edu | `skill/check-against-sources/SKILL.md`, through @skill-creator |
| Gemini | 18 or over, personal Google Accounts; work and school accounts "soon" | `skill/check-against-sources/SKILL.md` |
| A plan without skills | Everyone, through a project's instructions or one chat | `portable/paste-in.txt` |

"Who can add skills" is what each vendor's own page says, as of October 2026. These tools change often.

## Licence

Free, under the Neosaipien free licence, Version 2. Use it as much as you like, at work or at home, in any organisation. Change it to fit your team, and hand the changed version to people inside your own organisation, team, class or group, with the credit, the date and a line saying it was changed. Share unchanged copies with anyone, free of charge. Do not sell it, and do not publish a changed version outside your own group. Full terms in `LICENSE.txt`.

Neosaipien is not affiliated with Anthropic, OpenAI or Google. Product names belong to their owners.

Questions: samson@neosaipien.com, neosaipien.com

## Sources for "Why it matters now" (read 16 October 2026)

1. WorkLab, "2026 Work Trend Index report: Agents, human agency, and the opportunity for every organization", 5 May 2026: https://www.microsoft.com/en-us/worklab/work-trend-index/agents-human-agency-and-the-opportunity-for-every-organization
2. Anthropic, "Anthropic Economic Index report: Cadences", 26 June 2026: https://www.anthropic.com/research/economic-index-june-2026-report

## Sources for the setup steps (all read 16 October 2026)

- Anthropic, Skills overview: https://claude.com/docs/skills/overview.md
- Anthropic, Claude plans and pricing: https://claude.com/pricing
- Anthropic, What are projects: https://support.claude.com/en/articles/9517075-what-are-projects
- Anthropic, Connectors, skills, and plugins: https://claude.com/docs/extend/overview
- Anthropic, Change the model, effort, and thinking settings: https://support.claude.com/en/articles/8664678-change-the-model-effort-and-thinking-settings
- OpenAI, Skills and plugins: https://learn.chatgpt.com/docs/skills-and-plugins.md
- OpenAI, Build skills: https://developers.openai.com/plugins/build/skills.md
- OpenAI, ChatGPT release notes: https://help.openai.com/en/articles/6825453-chatgpt-release-notes
- OpenAI, ChatGPT plans and pricing: https://chatgpt.com/pricing/
- OpenAI, Projects in ChatGPT: https://help.openai.com/en/articles/10169521-projects-in-chatgpt
- OpenAI, GPT-6 and other models in ChatGPT: https://help.openai.com/en/articles/20001354-gpt-6-and-other-models-in-chatgpt
- Google, Create and manage skills for Gemini Apps: https://support.google.com/gemini/answer/17094296
- Google, About the transition from Gems to skills: https://support.google.com/gemini/answer/18560919
- Google, About Gemini models: https://support.google.com/gemini/answer/18648722
- Agent Skills overview: https://agentskills.io/
- Google, NotebookLM is now Gemini Notebook (16 Jul 2026): https://blog.google/innovation-and-ai/products/gemini-notebook/notebooklm-gemini-notebook/