We try to kill every idea.

The checks

Every idea is attacked on the same fronts, and dies on the first one where cited evidence is found against it. A listed pack is one where no hard gate produced that evidence.

These are common to every idea. Some face more: a small side-business idea is also tested on whether buyers are actively searching, whether the trend is still current, and whether its claims can be checked. Each pack page names the checks that idea faced and how many it cleared.

1. 1

   Real pain

   Is the pain real, or are we imagining it?

2. 2

   Lasting value

   Does the value last, or does it decay?

3. 3

   Room past the incumbents

   Have the big players already won?

4. 4

   A payer who can pay

   Is the buyer actually solvent?

5. 5

   A route to the buyer

   Can this reach a market at all?

6. 6

   No legal landmine

   Is there a regulatory path?

That is the short version. How it works is the long one: every check above, with a real idea it killed and the sourced argument that killed it.

The kill log

Most of the ideas the engine runs are killed. We do not hide that. Every kill is in the kill log, with the cited argument that killed it. The log is the receipt behind the catalogue; the catalogue is the survivors of the log.

The voice is source-or-die. Sourced, not sold. Refutational, not promotional. If a claim has no source, it is not in a pack. If an idea cannot survive the filter, it is not on the shelf.

What a pack actually is

A build spec, a go-to-market plan, an operations playbook and a QA report, with a citation behind every claim and a date stamped at publish. Yours outright: no login, no dashboard, no subscription, and plain text files you can open anywhere.

See one for free, no card, no email, on the sample page. The same rigour that produced the catalogue produced it.