GPSR disclosures for Shopify

Your suppliers send PDFs. Your products need disclosures.

Angaba extracts GPSR safety data from supplier documents — manufacturer, EU responsible person, warnings, identifiers — and writes it to Shopify's native disclosure field after you approve it. Every value shows the page it came from.

Built for the EU General Product Safety Regulation (GPSR), in force since 13 December 2024. Angaba is a data tool, not legal advice.

Launched on the Shopify App Store in September 2026 — by Gigliotti Software, the Lugano studio behind Rechna, Sichta and Anmelda.

Built for
  • GPSR (EU) 2023/988
  • shopify.disclosure
  • Shopify Sidekick
  • CE / EN 71

Why the native field changes things

Until mid-2026, compliance apps displayed safety data through its own theme widget or custom metafields — the data was locked to the app, and uninstalling broke the storefront. In June 2026 Shopify shipped a native disclosures field: disclosure data stored as first-party Shopify data, rendered by Shopify itself on the product page and in the cart of supported themes, in the Shop app and on AI shopping channels.

Angaba writes your GPSR data into that field. It stays your data, on your products, rendered by your theme — and it survives if you uninstall the app. Display is solved. What remains is getting correct data in, and that is what Angaba is for.

shopify.disclosure · list.disclosure_reference · rendered natively on supported themes, theme-block fallback for the rest

Two rules Angaba never breaks

Nothing is published without your approval

Extraction fills a review queue, not your storefront. You see every field next to the source document, edit or reject it, and only your explicit approval writes to the product. Every write lands in an append-only audit log with a gapless ANG-ID.

An empty field beats an invented one

Fields below the confidence threshold arrive empty and flagged — never guessed. Angaba will not invent an EU responsible person: if your documents don't name one, the scanner tells you what's missing and what an RP is, so you can obtain one.

How it works

Upload supplier documents

Declarations of conformity, manuals, spec sheets, test reports — the PDFs your suppliers already send you. Several PDFs per job, 25 MB for the job as a whole. PDF only for now: export scans or Word files to PDF first.

Angaba extracts and checks

AI reads the documents; deterministic validators check every value — formats, EU addresses, whether the document even matches your product. Each field carries the page and quote it came from.

You review and approve

Side by side with the source document. Approve, edit or reject per field. Documents for the wrong product are flagged whole, never partially applied.

Written to the native field

Approved data becomes a disclosure in Shopify's native field — rendered by your theme on the product page and in the cart, no widget. Shopify carries that field into the Shop app and AI shopping channels.

Angaba's compliance scanner dashboard: Ready, Pending and Missing disclosure counts per product, with a theme-compatibility check below.

What's in the box

Native disclosure writer

Creates the disclosure entries and writes shopify.disclosure per product — EU member-state jurisdictions handled correctly (there is no single "EU" code; Angaba knows that).

AI document extraction

Manufacturer, EU responsible person, CE marking, warnings, model/batch/GTIN — out of the PDFs, with per-field confidence and provenance.

Compliance scanner

14 deterministic checks per product: what's present, what's missing, what's out of sync — ready, pending or missing per SKU, with a plain-language fix for each finding.

Review queue with provenance

Every extracted value links to the page and sentence it came from. Checking a field takes seconds, not a re-read of the whole manual.

Theme block fallback

Theme without native disclosure support? One app block renders the same data — and removes itself automatically when the native section is present. Never twice.

Your data, portable

Disclosures live in your shop's own metaobjects and metafields, not in Angaba's database. Uninstall and everything you published stays.

Ask Sidekick

Angaba plugs its compliance data into Shopify's Sidekick assistant. Ask “which products aren't GPSR-ready?” or “did my supplier documents finish extracting?” right in the admin — answers come from the same deterministic checks, never guesses.

Bulk by CSV or identifier

Import your whole catalog's safety data from one CSV, or let one supplier document set cover many products — matched by exact SKU or GTIN, ambiguous matches held for you to decide, every write reviewed and audit-logged.

Two ways an app can store your disclosure data

In your shop, natively

Angaba writes into Shopify's own shopify.disclosure field, backed by metaobjects that live in your shop. It's your data, on your product. Uninstall Angaba and the disclosures stay exactly where Shopify put them — rendered by Shopify itself, no app required.

In the app's own database

Some compliance apps keep safety data in their own database or app-owned metafields, and render it through a theme widget or script tag they control. Uninstall that app and the widget — often the data with it — goes with it.

Ask any compliance app which of the two it writes to — the answer is in your Settings → Custom data → Products after one save.

Not for you if

  • You sell only outside the EU and Northern Ireland
  • You need US Prop 65 or CPSC disclosures — Shopify ships those natively
  • You need a warning/pictogram library, batch or serial-number tracking, or a generated Declaration of Conformity PDF
  • You want warnings drafted from product titles, without supplier documents to extract from

Built to refuse before it's wrong

A wrong EU responsible person or a missing warning is a liability, so Angaba's extraction is engineered around refusal. On our test set of 28 real supplier documents in five languages, every manufacturer value that cleared the confidence gate was correct: 31 of 31, name and address. No responsible-person value cleared the gate; two were sent to review and the remaining documents named none. All six documents with no product-safety content — three technical data sheets, an invoice, a safety-irrelevant sheet and an unreadable scan — were refused rather than mined for a plausible-looking name. Eleven fields in total went to review instead of being guessed. Harness run of 1 September 2026.

How we measure this →
Invoice instead of a DoC? Refused Document for a different product? Flagged, not applied Unreadable scan? Refused, not guessed Test lab is not the manufacturer

Where your data goes

Pricing

Free covers your first products and the native-field write — that part is never gated. Paid plans add scale and extraction. All billing runs through Shopify.

Free

$0/month
  • 10 products with disclosures
  • Manual data entry
  • Native disclosure field write
  • Theme block fallback
  • Scanner status for your first 10 products
  • 3 free AI extraction credits, once per shop
Install on Shopify

Grow

$19/month
  • Per-check fix guidance on every product
  • Validated bulk CSV import
  • Full audit trail of every write
  • Automatic re-scan when a product changes
  • Unlimited products, CSV export
  • 7-day free trial, once per shop
Install on Shopify

AI

$49/month
  • Everything in Grow
  • 100 document-credits per month
  • AI extraction with review queue
  • Per-field provenance
  • Credit packs $12 per 100, under a $60/month cap
Install on Shopify

1 credit = 1 extracted PDF (25 MB per job, all files together; up to 50 pages per document). Flagged or failed extractions don't consume credits. Credit packs bill as Shopify usage charges under a $60/month cap you approve once at subscription. 3 free extraction credits to try AI, once per shop; the AI plan has no further trial. Grow has a 7-day trial.

Questions merchants actually ask

What is GPSR and does it apply to my shop?

The EU General Product Safety Regulation (2023/988) applies since 13 December 2024 to most non-food consumer products sold to EU buyers — wherever the seller is based. It requires product listings to identify the manufacturer (name, postal address, electronic contact), an EU responsible person when the manufacturer is outside the EU, and applicable warnings and safety information. Whether it applies to your specific products is a legal question — Angaba manages the data, it doesn't give legal advice.

What is Shopify's native disclosures field?

Since June 2026, Shopify products carry a first-party shopify.disclosure field. Disclosures stored there are rendered by Shopify itself — on the product page and cart of supported themes — and, per Shopify, in the Shop app and AI shopping channels. The data belongs to your shop, not to an app. Angaba writes to this field; some compliance apps still render safety data through their own widgets.

What exactly does the AI extract?

Manufacturer name, postal address and contact; EU responsible person if the documents explicitly name one; CE marking and cited directives; consumer-facing warnings; model, batch and GTIN identifiers. Every value carries the page and verbatim quote it was read from, plus a confidence score. Low-confidence values arrive flagged for review, never silently filled. Input is PDF only in this version.

Can the AI publish something wrong to my store?

No. Extraction never writes to your products. It fills a review queue where you see each value beside its source; only your approval writes anything. Deterministic validators additionally downgrade values with format problems (invalid GTIN, non-EU responsible-person address) regardless of what the AI reports.

What happens to my disclosures if I uninstall Angaba?

They stay. Disclosure data is written to your shop's own metaobjects and Shopify's native field — it is your data, and Shopify keeps rendering it. Angaba's own records (audit log, extraction jobs, uploaded documents) are deleted per our data-retention rules after uninstall.

My supplier documents are in Chinese or Italian — does that work?

Yes. The extraction handles multilingual documents; our test set includes German, Italian, French and bilingual Chinese-English supplier documents. Warnings repeated in many languages are collapsed to one entry.

What does Angaba do when a document doesn't name an EU responsible person?

It leaves the field empty and the scanner fails that product with an explanation of what an EU responsible person is and that you need to appoint one when the manufacturer is outside the EU. Inventing one would expose you legally — so Angaba never does.

Does it work with my theme?

Shopify's current themes (Horizon, Dawn and family) render the native disclosures section automatically. For themes that don't, Angaba ships a theme app block that renders the same data and removes itself when the native section is present, so nothing renders twice.

Is Angaba legal advice?

No. Angaba is a data tool: it extracts, checks, and publishes product safety information under your review. Whether your products are compliant remains your responsibility as the merchant — the app is explicit about this in every relevant screen.

Is there a limit on document size or length?

One PDF. A job may carry several PDFs up to 25 MB combined, and each one costs a credit. Each document can run up to 50 pages — split a longer one into parts and upload the parts as separate documents.

What happens when I run out of credits?

Nothing gets queued or charged. Uploading more documents than you have credits for is refused up front, with a message showing how many you need. Buy a credit pack ($12 per 100, capped at $60/month) or wait for your included 100 to reset at the next 30-day period. Flagged or failed extractions never consumed credits in the first place.

Do you offer a free trial?

Grow includes a 7-day free trial, once per shop. AI itself has no subscription trial, but every shop also gets 3 free extraction credits to try AI, once per shop, on any plan — no subscription needed. Once those are used, extraction requires the AI plan.

Can I bulk-import disclosure data instead of using the AI?

Yes, on Grow and AI. Download the 15-column CSV template, fill it in per product handle, and preview every row before committing — nothing writes to a product until you approve the batch. A row that fails validation is reported with its line number; the rest of the batch is unaffected.

The disclosure data your products already need, in the field Shopify built for it.

Install on Shopify

Free plan available · Billing via Shopify · Your data stays yours