The automation hub, before it opens
Automation builds that save a company hours, checked before you see them.
We are building a hub with one focus: saving people time on their workload. It is a place to ask about automation, find builds that work, and list your own. The standard for the check is published, and every build goes through it before anyone sees it.
Three ways in.
Ask
If you want an answer
You have a slow, boring job and want it automated. Ask a question and get an answer, or find a build that already does it. Each build shows how it was checked, so you know whether it was run or only read. It costs you nothing, and an email address is all we need.
How a build is checked: automatic tests run on the file first, then an AI assistant reads it. We look over a sample of the results ourselves, and every refusal a builder disputes.
Build and list
If you build automations
n8n, Make, Zapier, Notion, spreadsheets, scripts: all welcome from day one. Listing is free. We offer an extra place to be found, a quality check you can show, a personal referral link and your own numbers. We do not promise you more sales, because we cannot know that yet.
The problem we are working on is visibility.
- 0
- On n8n's public library, paid templates created in the last 90 days have a median of 0 views (read by Fractional HQ, 10 Nov 2026).
- 292 of 3,113
- n8n lists 3,113 template creators and 292 of them sell (read by Fractional HQ, 11 Nov 2026).
- 30 of 59
- Review is slow too: of the 59 review verdict emails n8n has sent us, 30 were refusals or change requests, most often "too basic" or sticky-note and layout problems, and each cost us two to three days.
n8n's public data shows a creator views only.
So the hub gives you a clear standard up front, a recorded check, and a page of your own.
Connect an agent
If you connect an agent
Your agent can ask questions and search the library through a connection we run. It gets answers, summaries and links. It does not get a listing's file. Connecting costs nothing.
The quality line
- File and guide read
- Run on sample data: FHQ Tested
We only show builds that have been checked, and we say how. Each listing shows the level its check reached: file and guide read, or also run on Fractional HQ's test accounts with sample data, which carries the mark "FHQ Tested". If a build does not pass, you will not see it listed at all.
A check gives a sensible estimate of quality and stops short of a full audit. We confirm that the build opens cleanly, matches what its listing says when read, and comes with a guide you can follow. A build with the mark was also run on sample data. We do not test your own data, your account limits or future changes to the tool, and we do not promise it never fails.
A listing can give you up to three options: get the build as it is, ask the builder to set it up for you if they offer that, or ask Fractional HQ to set it up.
Join the list
What happens next
A confirmation email arrives within a few minutes. Open its link and press the button to confirm your place.
Next comes your personal link. Share it if you want to: when a friend joins through it and confirms, that counts for you. We never ask for your friends' email addresses.
Last, you get your place in the list, and we tell you the number.
The countdown
The countdown is for launch day only. Before it starts, the library of verified fixes and the agent connection are already open, so you can use something real while you wait. We will not start a countdown until there is a date we can keep.
Questions
Can an agent copy my build?
Not through our connection. It returns answers and short summaries, never the file. The file is not on the public page. One honest limit: anything shown on a public page can be read by a person or an agent who visits it, so we keep the file off the page.
What does it cost?
Joining, asking, listing and the check are free. Later we may sell promoted placement. If we do, it will be labelled as an advert and kept apart from checked results.
Who checks the builds, and how?
Automatic tests run on the file first. Then an AI assistant reads the file and the guide, using the published levels. A build is run on sample data only when its builder asks for a run, and only for a tool we can run, which is n8n today. We look over a sample of the results ourselves, and every refusal a builder disputes. The check is proportionate and stops short of a full audit, and each listing says what was done.
Which tools?
Any tool, from day one, whether it is n8n, a spreadsheet or a script. The level a build can reach depends on whether we can run that tool, and each listing shows its level.
What if nobody comes?
Then we will say so, with the numbers. We set a floor for 4 December: 150 confirmed sign-ups, 30 of them builders, and 5 builders from outside with a build that passed the check. If we reach it, the hub opens on 8 December and we make some noise about it. If we fall short, it opens quietly on the same day, as long as every part has been tested live. You will be told by email either way, and you can leave the list at any time.