How we curate

How an entry gets here

1. Discovery: we look through X, Product Hunt, GitHub, Reddit, Hacker News and newsletters, and readers submit links.
2. Drafting: each lead is written up in our standard format: what was done, how it went, the prompt, business fit and the source. Research bots can prepare these drafts, and they go into a review queue.
3. Review: an editor opens the source, checks the claims, sets the evidence grade, and rewrites anything that copies the original too closely.
4. Publishing: only approved entries reach the site. Nothing publishes automatically.

Evidence grades

• AIndependently verified. The outcome is confirmed by data or by someone not involved in the work.
• BFirst-hand account. The person who did the work describes the setup and results in detail.
• CVendor or secondhand. A tool maker's case study, a promotional demo, or someone else's write-up.
• DAnecdotal. A brief, early or in-progress claim. Treat it as an idea to test, not proof.

A grade describes the evidence, not the idea. A D-grade entry can still be worth trying; it just hasn’t been proven.

Editorial standards

• Business and entrepreneur use cases only.
• Every entry links to its source. If we couldn’t open the source ourselves, the entry says so.
• Results are stated as the source states them, with limitations kept in.
• Prompts are written or adapted by us unless marked otherwise; we link to the original instead of copying long passages.
• “Business fit” notes are our editorial suggestions, even when the source didn’t say who it’s for.
• Every use case and tool page opens with “Explain it simply”: two to four plain sentences for readers new to AI tools.
• When the source is the tool maker’s own case study, the page says so at the top.
• Every use case ends with “Similar tools”: other tools that do the same job, each with one line on how it differs, taken from the tool’s page or a linked source. Our editors choose them; they’re never paid for.
• Income Automations show income exactly as the person stated it. Any claim we haven’t independently verified carries a notice at the top of the page.
• Setup guides are written by our editors from the linked sources and link to the real examples they’re built from.

How we make money

Three ways, all labeled: weekly sponsors (the block at the bottom of each page and a line in the newsletter), featured tool listings (marked “Featured” in a verdict list), and affiliate links on some tool pages (marked “Affiliate link”). None of them changes a verdict, a ranking or an evidence grade. Sponsor a week.

Privacy

This site doesn’t run analytics or advertising scripts. The newsletter signup form at the bottom of each page comes from Kit, our email provider. When the form loads, it tells Kit the form was shown, and Kit keeps a random ID in your browser’s storage to count those views. If you subscribe, Kit stores your email address, and we use it only to send the weekly newsletter; unsubscribe from any issue. Use case submissions and sponsor bookings are stored by our host, Netlify, and read by our editors.

Corrections

Every entry has a “Suggest a correction” link. We fix verified mistakes and update the review date.

Open data

Everything published here is available as JSON at /data/index.json, with schemas at /schemas/. Agents and readers are welcome to use it with a link back.