How a verdict is made

This page is generated from the same rubric file my scorer reads, so whatever it says here is what got applied.

Principles

The bar is local, and it doesn’t move I compare projects to what gets built here, not to global startups. The rubric is fixed though, with written anchors for every band. On a curve your score would drop the day someone else ships something great, so here it only moves when your project does. That also means nothing has to reach Moat.

Engagement is never an input Views, reactions, subscriber counts and shares don’t touch any score. I only look at whether the thing works and whether it’s set up to last.

Every point has a receipt Scores come from the creator’s own public posts and the links in them. Every page the crawler checks gets saved with the date, so a receipt still holds up after the site goes down.

What counts as a project

Anything with a link people can go to and get value from: an app, a site, a bot, an API, a library. Experiments, demos and tutorials without one don’t count. Something informal becomes a project once its creator is still posting about it 30 days later. A project built across several channels gets one page, linked from each of them.

The scale

Five measures of 100 points each add up to a score out of 500, which lands in one of three bands. Moat is narrow on purpose.

Bloat
0

Float
300

Moat
400

500

Bloat: 0–299
Not much here works or looks likely to last, yet.
Float: 300–399
Real and alive, but nothing stops a copy yet.
Moat: 400–500
Real, and hard to copy. Rare on purpose.

100 points · judged against anchors

Viability

Is there a real problem, someone who would pay, and a market this builder can actually reach?

0–20No clear problem or user. A toy, a practice clone, or a solution looking for a problem.
21–40A real problem, but no plausible path to anyone paying, or a market well out of this builder's reach.
41–60Plausible users and a believable way to charge them. Not yet shown to work.
61–80People use it and pay (or clearly would), and the model fits the local market: how people pay, what they can afford, the language they use.
81–100A working business: paying customers, a model that already sustains it, and room to grow.

Not countedAudience size. Views, reactions, subscribers and shares never count, since none of them show that anyone needs the thing.

100 points · judged against anchors

Moat

What stops someone copying it?

0–20A thin wrapper or a clone that anyone here could rebuild in a weekend.
21–40The only edge is execution, or being first.
41–60Local depth that takes real effort to copy: payment integrations, Amharic, regulation, partnerships, operational know-how.
61–80Something compounding: proprietary data, network effects, owned distribution, or hard technology.
81–100Defended on several fronts, and hard to displace even for a funded competitor.

100 points · a checklist

Momentum

Did the updates keep coming?

Sustained shipping60 pts
How long the project has kept being worked on: points start after one month and reach the full 60 at six months.
Read off the timeline
Consistency40 pts
The share of months in that span with at least one update. Steady beats bursty.
Read off the timeline

Not countedHow many people saw an update. Activity is a creator's own posts about the project, its repo commits and store updates. After 60 quiet days the total starts to decay; after 90 the project is Dormant.

100 points · a checklist

Infrastructure

Does a real, working product exist?

It's live20 pts
The project's link works: the site responds, the store listing exists, or the bot answers.
Checked on the page
HTTPS5 pts
Served over HTTPS (granted for store apps and bots).
Checked on the page
Real product35 pts
Real functionality beyond a landing page or waitlist. 0 nothing usable · 15 landing page or demo · 25 working but thin · 35 a full working product.
Judged against the rubric
Own home10 pts
Its own domain or a store listing, not a free subdomain or link-in-bio page (a bot alone earns 5).
Checked on the page
Maintained15 pts
Something shipped in the last 90 days: a store update, a repo push, or an update post.
Checked on the page
Operations15 pts
Signs of real operations: accounts and a backend, docs, an API, a status page, more than one platform.
Judged against the rubric

100 points · a checklist

Sustainability

Is it being set up to last?

Distribution30 pts
How people get it. 30 a listing on an official store or registry (Play Store, App Store, Chrome Web Store, VS Code Marketplace, Raycast Store, npm, PyPI…) · 25 a web app on its own domain · 20 a Telegram bot or mini app · 12 a site on a free subdomain · 10 versioned downloadable releases (e.g. GitHub releases with an installer) · 5 an APK file passed around · 0 nothing.
Checked on the page
Payments15 pts
A checkout wired in: Chapa, SantimPay, ArifPay, Stripe, PayPal and the like. Naming a bank or wallet on the page doesn't count.
Checked on the page
Pricing10 pts
A pricing page or stated prices.
Checked on the page
Revenue5 pts
The creator's posts show paying customers or revenue.
Judged against the rubric
Privacy policy7 pts
A privacy policy page (or one on the store listing).
Checked on the page
Terms5 pts
Terms of service.
Checked on the page
Support8 pts
A way for users to reach the people behind it: email, a contact page, a support bot or group.
Checked on the page
Commitment20 pts
A team forming, a roadmap kept, a dedicated channel or community for the project, funding or an accelerator.
Judged against the rubric

Not countedWhether there's a registered company behind it. Builders here rarely incorporate, and it says little about whether a product will last.

When there’s no score

Under review: Too early to call. It showed up less than 30 days ago, or none of its links checked out yet.
Dormant: No updates in 90+ days, so its Momentum is decaying.
Acquired: Another company bought it, which answers everything the rubric asks. It gets full marks.
Closed: The creator shut it down. The score stays where it was.

Momentum starts decaying after 60 quiet days, and at 90 the project gets marked Dormant. A Closed project keeps the score it had when it shut down.

Who checks this

Claude does the reading. It finds the projects in each channel’s posts and scores them against the rubric, while the crawler checks their pages. A second Claude pass reviews each project and verdict before it goes up, and it only passes a verdict that cites evidence for every judged score. I can overrule any of it.

If something shouldn’t be here, report it. Once three different people report the same project it comes down straight away and stays down until I’ve looked at it. Scores get re-run every month. If it’s your channel you can claim it, and then reply under any verdict ( the reply shows up next to it ) or ask me for a re-review. How claiming works.