The golden thread for fire doors: what to keep with the door

Ask a straightforward question about one fire door in one block — when was it last checked, by whom, what did they find, and what happened next — and the answer usually depends on who still works for you. The check happened. The record exists. It is in a spreadsheet a contractor keeps, a PDF in somebody’s inbox, or a photo on a phone that left the company two years ago.

That is the problem the golden thread is meant to solve, and it is worth being precise about what it is and where it actually bites.

Where the phrase comes from

“Golden thread” entered the language through Dame Judith Hackitt’s 2018 review Building a Safer Future, which described building safety information that is created once, kept accurate, and handed on intact through a building’s whole life — rather than reassembled from scratch each time somebody asks.

The Building Safety Act 2022 turned that principle into statutory duties for higher-risk buildings — broadly, buildings of at least 18 metres or seven storeys containing at least two residential units. If you are responsible for one of those, the golden thread is a legal obligation with its own regime, and this guide is not a substitute for reading it.

If you are not, the principle is still the thing that answers “show me the record” — from a resident, an insurer, a new contractor, or a fire and rescue service. Nobody asks whether your evidence was legally required. They ask whether you have it.

The test a record has to pass

One question, about one door, answered today, without ringing anybody:

• What is it, and where exactly is it?
• When was it last checked, and by whom?
• What were they checking against?
• What did they find?
• What happened to what they found?

If any of those needs a phone call to a contractor, you do not have a record of that door. You have a contractor who has one.

What to keep with the door

Seven things. In roughly this order of usefulness.

An identity that stays put. A reference that is physically on the door, not a row number in whatever system happens to be current. Everything else hangs off this.

Where it is, precisely enough to find it again. Building, block, floor, and the specific opening — a stair door on the fourth floor is not a location if there are three of them. A pin on a floor plan removes the ambiguity entirely.

What the door is. Type, the rating as it has been recorded, hardware, glazing, signage. Record what you know and note what you are unsure of. A guessed fire rating written down as fact is worse than a blank.

Installation and product evidence, where it exists. Certificates, data sheets, the installer, the date. For a door fitted before you started keeping records this will be empty, and an honest empty is fine — an invented one is not.

Every check, in full. The date, the person, the form they used and which version of it, the outcome, and photographs. “Which version of the form” sounds pedantic until the form changes and you need to know what the 2024 checks actually asked.

Defects, and their endings. A record that captures problems but not their resolution proves you knew and does not prove you acted. The close-out matters more than the finding.

The end of the door’s life. Replacement or removal, dated. A record that stops without saying why looks like a record somebody abandoned.

Why it has to belong to the door

Contracts change hands. When they do, the paperwork goes with the outgoing contractor and the incoming one starts from nothing — re-surveying doors that were surveyed eighteen months ago, and producing a history that begins on the day they arrived.

That is not a filing failure. It is what happens when the record is attached to the supplier rather than to the asset. The door has not changed. The building has not changed. Only the company holding the folder has.

So the useful move is not a better folder. It is putting the identity on the door itself, and letting whoever is holding a phone in front of it read what is already there.

What a tag actually does

This is where MyQTAG’s argument sits, so here is the mechanism rather than the claim.

Every asset gets a durable QTAG label carrying a reference and a QR code. The code resolves to that tag’s page, and it opens in any phone camera — no app, no account needed to scan it.

The tag is the only thing shared between organisations. Each team keeps its own site records, its own copy of the asset and its own inspections, exactly as if the other team did not exist. What crosses the boundary is the tag’s history — so a contractor who arrives in 2026 and scans a door tagged in 2024 sees the work that came before them, and the building manager sees both.

What crosses is the work, not the worker. A cross-team reader sees the performing organisation, the form, the reference, the dates, the status and outcome, the asset and its location, and the photographs. They do not see who submitted it, who approved it, the reassignment history, or anybody’s contact details. That is a deliberate line: your contractor’s staffing is their business, and the evidence about the door is yours.

Three lifecycle cases are worth knowing, because they are the ones that break naive systems:

• A scuffed label is replaced, and the new tag carries every organisation’s link across. The old reference — the one printed on paperwork you have already issued — still resolves, forward, to the tag now on the door.
• A replaced door is a clean break: a new asset, a new tag, and an empty history. That is correct, not a gap. The old door’s record closes rather than being quietly inherited by a different piece of joinery.
• A removed asset closes its record and flags every other organisation linked to that tag, so their register shows the door was reported removed and by whom. Nothing is auto-archived on somebody else’s behalf.

What this does not do

Three honest limits, because a guide that only lists strengths is an advert.

A tag is not a fire risk assessment, and a register of checks is not one either. It is evidence that the checks your assessment calls for happened.

Nothing reconciles disagreements. If your record says a door is FD30 and your contractor’s says FD60, both stand. Each organisation’s records are its own compliance trail, and silently overwriting one with the other would be worse than the disagreement. What the shared history gives you is the ability to see the disagreement, which is the part you currently cannot do at all.

There is no single master record of the asset. The tag is the convergence point, not a canonical file that everyone edits. That is a deliberate trade: it keeps each organisation’s data its own, and it is why there is no permissions matrix to administer.

Where to start

You do not need to tag an estate to find out whether this works. Take one block, tag its common-area doors, and run one round of checks through it. The question to ask at the end is not whether the app was pleasant. It is whether somebody who has never been to that building could answer the five questions above from the record alone.

If you would like to try that on a building of yours, book twenty minutes and we will walk through it with your own floor plan.