Preview build for review — Saturday 14 November 2026 · Launch catalogue: 10 manuals · Design unchanged · Import is upload & source-URL only · Full category tree + original guides · Live site: custom WordPress theme + HandbookNest Core — zero extra plugins
HandbookNestEvery handbook. One nest.

Every handbook.
One nest.

User manuals, installation guides, service documents and fixes for the devices in your home — searchable by brand and model number, readable in the browser, and downloadable as a clean HandbookNest-branded PDF.

—brands in the seed catalogue
—handbooks ready to open
100%free to read & download
Illustration of an open handbook resting in a woven twig nest with amber leaves — the HandbookNest logo artwork
Fig. 01 — The nest mark: an open handbook, cradled. HandbookNest identity artwork.

Browse by brand

Each brand gets its own hub page — /brand/samsung/ — with every model and document type in one place.

Browse by category — the full tree

The complete factual category tree for the library: brands and product types are the shared labels every manuals library uses, organised here in HandbookNest's own structure. The launch catalogue fills the first branches with its 10 manuals; the remaining branches are ready in the taxonomy for the next imports. Don’t know the model number? Start from the appliance or device and narrow down from there.

The library

Search matches brand, model number, document title and the extracted text of each handbook — the same behaviour the dedicated search index will provide at launch.

How a handbook reaches you

Three steps for readers. Behind them sits the ingestion pipeline in the blueprint below.

1

Find it by model number

Search is tuned for the strings people actually type — model numbers, FCC IDs and error codes — not just product names.

2

Read it in the browser

Every handbook is also published as a fast HTML page with a contents list, so the answer is one tap away without opening a 200-page PDF.

3

Download the branded PDF

The PDF carries a HandbookNest cover, footer and page numbers. Try it — the download buttons in this preview generate a real sample PDF in your browser.

Quick answers — written to be quoted

This is the AEO layer in action: short, direct, self-contained answers on the page, backed by matching FAQPage schema in the head, so Google featured snippets and AI answer engines (ChatGPT, Perplexity, Gemini, Copilot) can lift the answer and cite HandbookNest as the source. Every /fix/ and handbook page follows the same pattern: answer first, detail after.

How do I find a manual by model number?

Type the exact model number from the rating plate — for example SM-X820 or SHSM63W55N — into search above. The model page gives the user manual, installation guide and any fix guides for that exact model, free to read and download.

What does Bosch dishwasher error E24 mean?

E24 means the dishwasher is not draining properly. Check the filter and drain pump cover for blockages, check the drain hose is not kinked or blocked at the air gap, then reset the control panel. The full step-by-step is in the Bosch SHSM63W55N guide in the library.

Are HandbookNest manuals free?

Yes. Every handbook is free to read in the browser and free to download as a HandbookNest-branded PDF. No account, no paywall.

Where do the manuals come from?

From the manufacturer's own support site or another authorised source. Each record stores its source URL and rights basis, the manufacturer's notices stay in the document, and the pipeline only adds a cover, footer band and metadata — the manual body is never altered or rewritten.

Guides & articles — written by HandbookNest

Original HandbookNest writing: practical explainers that help you find the right manual, understand an error code and read the label on your device. Written in-house from the manuals themselves and everyday repair experience — not taken from any other site.

Getting started · HandbookNest guide

How to find your model number — and why it matters more than the product name

“Samsung fridge” or “the big Dyson” will not find your manual. Manufacturers publish a different document for almost every model variant, and two machines that look identical on the shelf can need different parts, filters and reset steps. The model number is the one string that picks the exact book.

Where to look:

  • Dishwashers & washers: on the door edge or just inside the door frame.
  • Fridges: inside the fresh-food compartment, on the side wall.
  • TVs & phones: on the back panel, and again in Settings → About.
  • Small appliances: on the base — it is often moulded, not printed, so tilt it to the light.

Copy it exactly, including dashes and trailing letters — SHSM63W55N and SHSM63W55N/01 can be different revisions. Then paste it into HandbookNest search: one model, one page, the right manual.

4 min readPairs with: any brand hub
Error codes · HandbookNest guide

Bosch dishwasher error E24, in plain English

Short answer: E24 means the dishwasher tried to drain and the water did not leave. It is a drainage fault, not a dead pump by default — in most homes it is something simple blocking the way out.

Work through it in this order, cheapest first:

  • Filter first. Twist out the cylinder filter in the base, rinse it under the tap and check underneath for glass, labels or bone fragments.
  • Pump cover next. In the sump beside the filter there is a small cover over the drain pump impeller. Unclip it (see the manual for your exact model) and clear anything wrapped around the impeller, then click the cover firmly back.
  • Then the hose. A kinked drain hose, or a blocked connection where the hose meets the sink trap or air gap, will trigger E24 even with a clean machine.
  • Reset last. With the blockage cleared, cancel the programme to force a drain cycle. If E24 returns on an empty, clean machine, that is the point to call service — the pump itself may have failed.

The full illustrated steps for the Serie 6 are in the Bosch SHSM63W55N guide in the library.

5 min readPairs with: Bosch SHSM63W55N guide
Know your device · HandbookNest guide

How to read a rating plate: model, serial, voltage and those small certification marks

Every appliance carries a rating plate — the small silver or white label that looks like fine print and is actually the device's identity card. Learning to read it takes two minutes and saves a lot of wrong manuals and wrong parts.

  • Model number — the line you search by. Sometimes labelled Model, Type or E-Nr. depending on the maker.
  • Serial number — unique to your unit. You rarely need it for a manual, but warranty and recall checks ask for it, so photograph the plate once and keep it.
  • Voltage & frequency — e.g. 120 V ~ 60 Hz or 220–240 V ~ 50 Hz. This is why the same model name can have region-specific manuals: the wiring inside really is different.
  • Certification marks — UL, ETL, CE, UKCA and the FCC ID on anything with a radio. An FCC ID on the plate is also a lookup key: certification records link back to the exact model they belong to.

Rule of thumb: if a support page asks you to “confirm your model”, it means this plate — not the marketing name on the front and not the box in the loft.

4 min readPairs with: certification records

Admin panel — import a manual in under a minute

This is a working model of the HandbookNest Core admin. Watch the narrated walkthrough, then try both routes yourself — Import from source URL for a direct PDF link from the manufacturer or another authorised source, or Upload manual for a PDF you already have. That is the whole import system: no other route, nothing fetched from anywhere else. Everything lands in the one review queue below, and nothing publishes without review.

Demo video: the admin panel, start to finish

Voice-over walkthrough · about 1½ minutes · the on-screen panel highlights each step as it is narrated · Download the MP4 with voice-over

Read the transcript

Welcome to the HandbookNest admin walkthrough. In about a minute, you will see exactly how a manual gets from a manufacturer website, or from your own computer, onto a live HandbookNest page. Start on the admin dashboard. You have two ways in, and both end in the same review queue. First, Import from source URL. Paste a direct link to a manual on the manufacturer's own website, then add the brand, the model, and the document type. Before you can queue it, you record the source and your rights basis: manufacturer or otherwise authorised sources only. The original copyright notice stays inside the document. HandbookNest branding adds our cover, footer, and page numbers, and the handbook page adds its own summary, contents, and specifications written in our own words. Branding never changes who owns a manual, and it is never a way around copyright. Second, Upload manual. Choose a PDF from your computer, add the same brand and model details, and it joins the same queue. In the review queue, every item shows its source, its rights basis, and its pipeline status: duplicate check by file hash, text extraction, branding, and a human review before anything is published. Nothing goes live automatically. Finally, open SEO and Indexing. Connect Google Search Console to verify the site and submit the sitemaps, and add your GA4 measurement ID so views and downloads are measured. GA4 measures traffic; Search Console is what helps Google find and index the pages. That is the whole flow: import, review, publish, and watch it get indexed.

Import from source URL

One press pulls a manual straight from the manufacturer's own website into the pipeline — branded, summarised in our own words, and published as a HandbookNest page.

The source URL and rights basis are stored with the record, so every published page can show exactly where its manual came from.

Upload a manual yourself

Already have the PDF — from a product box, an email from the maker, or your own archive? Upload it directly; it goes through the exact same pipeline.

In this preview the file is checked and fingerprinted in your browser only — on the live build it is stored in Cloudflare R2.

How HandbookNest stays on the right side of copyright — by design, not by disguise. Branding and restyling do not change a manual's copyright status, and they don't remove Google duplicate-content or rights-holder issues — so the system never relies on them, and nothing here is built to hide copying from Google. Every file comes in through just two doors: Import from source URL (a direct PDF link from the manufacturer or another authorised source) or Upload manual (a PDF you supply yourself). There is no other intake. The manual's content then stays exactly as published: the pipeline only adds the HandbookNest cover, footer band and metadata, and never alters or disguises the body. Source and rights basis are logged per record, the manufacturer's notice stays in the document, HandbookNest adds value in its own words (summary, contents, specs, troubleshooting), and a published takedown / DMCA route handles any rights-holder request. If a source isn't authorised, it doesn't go in the queue.

Review queue

One queue for both routes. Demo queue for this walkthrough — items stay in this tab.

0 waiting

The queue is empty. Import a manual from a source URL, or upload a PDF, and it will appear here with its provenance and pipeline steps.

SEO, AEO & GEO — built in, on-page and off-page

Yes — the design stays exactly as you approved it: same Fraunces + Inter type, same warm bark / amber / clay palette, same search-first, text-first simplicity with no clutter. What sits underneath is the full visibility stack: classic SEO, AEO (answer engines & featured snippets) and GEO (generative engines — ChatGPT, Perplexity, Gemini, Copilot), an AI-crawler-friendly robots.txt, and llms.txt — all generated by HandbookNest Core, with direct connections to Google Search Console and GA4. No SEO plugin, no extra plugins at all.

Connect Google — Search Console & GA4

On the live build these two fields live in WordPress under HandbookNest → SEO & Indexing. Paste your values here to see exactly what the connection does.

Not connected

Connecting writes the google-site-verification meta tag into every page head (the placeholder tag is already in this page's head), then submits the sitemap index to Search Console. On the live build, Core also syncs indexing status back through the Search Console API, so the dashboard shows which handbook pages Google has actually indexed.

Not connected

Connecting loads GA4 with three custom events wired up: manual_view, pdf_download and manual_search (with the model number as a parameter). One honest note: GA4 measures traffic — it does not help indexing. Indexing is Search Console's job.

Indexing Control Center — launch snapshot

What the in-admin Indexing Control Center reports once GSC is connected. Figures below are this preview's launch state: the 10-manual catalogue, sitemaps generated from it, ready to submit.

Pages published10
In sitemap index10
Sitemap files3
noindex by ruleSearch · thin combos
  • Sitemap index at /sitemap_index.xml, sharded by brand and content type — generated from the same catalogue as this page.
  • robots.txt allows every major AI crawler by name (full list in the AEO & GEO panel below), allows the library, blocks only search-result and admin paths, and points crawlers at the sitemap index and llms.txt. Download the real file below and check it yourself.
  • Built only on supported Search Console APIs: the Sitemaps API submits sitemaps and reads their status; the URL Inspection API syncs each page's real index status (quota-aware, batched); the Search Analytics API pulls clicks and impressions into the same dashboard.
  • Publish hooks + worklist: every publish or update runs pre-flight checks (canonical, robots, schema, sitemap membership) first, then the page enters an index-status worklist in admin until Search Console confirms it. IndexNow pings Bing and Yandex on the same hook.
  • Indexing API — deliberately not used: Google documents its Indexing API as only for job-posting and livestream pages, with spam detection and possible access revocation for misuse — so it plays no part here. Google finds pages through sitemaps, internal linking and IndexNow.

All three files are generated in your browser from the live 10-manual catalogue above — open the sitemap and you'll see the real /brand/ and /guide/ URLs it would submit to Search Console, and the robots.txt carries the explicit AI-crawler allow-list.

On-page SEO — shipped in the theme & Core plugin

Every page, by default

  • Unique title & meta description per handbook, generated from brand + model + document type — try this page's head: canonical, robots, Open Graph and Twitter cards are all in.
  • One canonical URL per handbook — the HTML page. The branded PDF is a download, never a competing duplicate.
  • Structured data: BreadcrumbList on every page (live in this page's head), WebSite + SearchAction for sitelinks search, and per-handbook Product / TechArticle + FAQ schema on /fix/ guides — sample below.
  • Internal linking: brand hubs ↔ handbook pages ↔ related models ↔ fix guides, precomputed in Core so pages link richly without slow queries.
  • Speed: local fonts, no page builder, Cloudflare full-page cache — Core Web Vitals treated as a ranking input, not a report.
Per-handbook JSON-LD (emitted by Core on /guide/ pages):
{
  "@context": "https://schema.org",
  "@type": "TechArticle",
  "headline": "Bosch SHSM63W55N Dishwasher — E24 Guide",
  "url": "https://handbooknest.com/guide/bosch/shsm63w55n/",
  "isPartOf": { "@type": "WebSite", "name": "HandbookNest" },
  "about": { "@type": "Product", "name": "Bosch SHSM63W55N",
             "brand": { "@type": "Brand", "name": "Bosch" } }
}

Off-page SEO — the authority plan

Off-page can't be coded into a theme, so it ships as a working programme with a tracker in the admin. The rules: earn links with useful content, never buy them, and keep every claim about a manual accurate.

  • Fix-guide outreach: the /fix/ error-code guides (like the Bosch E24 page) are the linkable assets — pitched to appliance-repair blogs, forums and community wikis as the citable source.
  • Brand & community presence: helpful, attributed answers in owner forums and Q&A threads that link to the exact model page — no spam, no link schemes.
  • Digital PR: data stories from the library (most-searched error codes, manuals people can't find) offered to home-tech press.
  • Profiles & citations: consistent HandbookNest name, URL and description on the directories and social profiles that matter.

Backlink & outreach tracker

AEO & GEO — found and cited by answer & AI engines

AEO — answer engine optimisation

  • Answer-first pages: every handbook and /fix/ page opens with a direct, self-contained answer (what it is, what the error means, the fix) before the detail — the format featured snippets and voice assistants lift. See the Quick answers section above: it is live AEO content, not a mock-up.
  • FAQPage schema on the homepage and fix guides (live in this page's head), plus HowTo-style step structure on troubleshooting guides and question-form headings that match how people actually ask.
  • Exact-match entities: brand + model + error code in the title, H1 and first paragraph, so “Bosch E24” style queries land on one unambiguous page.

GEO — generative engine optimisation

  • AI crawlers explicitly allowed in robots.txt — no blanket blocks, no training/search exclusion. If an engine can't crawl it, it can't cite it:
GPTBot ✓ AllowChatGPT-User ✓OAI-SearchBot ✓ClaudeBot ✓Claude-User ✓Claude-SearchBot ✓PerplexityBot ✓Perplexity-User ✓Google-Extended ✓Applebot-Extended ✓CCBot ✓Meta-ExternalAgent ✓Meta-ExternalFetcher ✓Bytespider ✓cohere-ai ✓YouBot ✓Amazonbot ✓DuckAssistBot ✓Googlebot · Bingbot ✓
  • /llms.txt published — a plain-text map of the library (brands, models, fix guides) written for LLMs to read and cite. Download the generated version from the panel above.
  • Citable, verifiable facts: specs, page counts and source provenance stated plainly on every page, with the manufacturer source named — engines cite pages they can verify, and consistent HandbookNest entity data (Organization + WebSite schema, same name / URL everywhere) makes the citation point here.
  • Clean, crawlable HTML: no JavaScript wall, no login, fast server-rendered pages — the same simplicity that helps Google helps every AI crawler.

The build blueprint

The professional WordPress approach, designed before a single template is written: presentation in the theme, everything that matters in a core plugin, and PDFs processed by a separate pipeline. This is the plan the live build follows.

Custom WordPress theme — presentation only

Homepage, search, brand hubs, handbook pages, device pages and troubleshooting guides. Local Fraunces + Inter fonts, plain CSS, no page builder — so pages stay fast and a future redesign touches nothing but looks.

  • Templates: front-page, brand hub, single-handbook, fix guide
  • Reusable blocks for contents lists, spec tables and download panels
  • Responsive design in the warm bark / amber / clay identity

HandbookNest Core plugin — the data layer

Everything structural lives here, separate from the theme: post types, taxonomies, rewrite rules, metadata, schema and sitemaps. Redesign the theme tomorrow and the library is untouched.

  • Custom post type handbook; taxonomies for brand, product type, document type
  • Precomputed related-manual links — no expensive queries at render time
  • Schema output: breadcrumbs, product/model data, FAQ

Custom tables where it counts

Postmeta collapses under heavy search. Fields that get searched — model numbers, variants, certifications, extracted page text — live in dedicated tables designed for it.

  • hn_manuals · hn_models (variants) · hn_documents
  • hn_pages — extracted text per page, feeds search + HTML view
  • hn_certifications · hn_import_jobs — FCC records & pipeline runs

Cloudflare R2 — the PDF store

Originals and branded PDFs live in R2, not the WordPress uploads folder. Free egress keeps a download-heavy library cheap: the free tier covers roughly the first 1,000 manuals stored as original + branded copies.

  • Free tier: 10 GB storage, 10M reads / month, $0 egress
  • ~10,000 manuals (≈100 GB) ≈ $1.35 / month storage
  • ~100,000 manuals (≈1 TB) ≈ $14.85 / month storage

Permanent URL structure — decided before development

PatternExamplePurpose
/brand/{brand}//brand/samsung/

Brand hub: every model, category and document type for one maker. The page that ranks for “Samsung manuals”.

/guide/{brand}/{model}//guide/samsung/sm-x820/

The handbook page itself. This HTML page is what Google indexes; the branded PDF is its download.

/fix/{brand}/{topic}//fix/bosch/dishwasher-e24-error/

Troubleshooting guides built from manual content — error codes, resets, maintenance.

/cert/{id}//cert/fcc-a3lsmx820/

Certification records (FCC and equivalents) linked from the model they belong to.

The PDF pipeline

Stage 1

Ingest

Manufacturer PDFs arrive via Import from source URL or Upload manual and enter a queue — processing never runs inside an admin page load.

Stage 2

Dedupe

File hashes and model matching catch duplicates and near-duplicates before they can split rankings across two pages.

Stage 3

Extract

Text, contents structure and images are pulled out per page; scanned manuals fall back to OCR.

Stage 4

Brand

A HandbookNest cover, footer band, page numbers and metadata are applied to produce the branded PDF. The manual's own pages are never altered — the same treatment the download button in this preview demonstrates.

Stage 5

Store

Original and branded files go to Cloudflare R2; WordPress keeps only metadata and links.

Stage 6

Publish & index

The handbook record lands in WordPress, the search index updates, and the page joins its sharded sitemap.

Backend processing — Python worker built for one Hostinger KVM 4

Two routes, one queue, one worker

  • Import from source URL — you paste a direct PDF link from the manufacturer's support site or another authorised source (including primary public databases such as FCC / Energy Star filings where they are the original publisher). You record brand, model, document type and rights basis; the worker then fetches that one file from the source you gave it. Nothing else is fetched, and no other site is involved.
  • Upload manual — you supply the PDF file directly from your computer or archive. It enters exactly the same queue with the same provenance record.
  • Shared pipeline for both — dedupe, text extraction, branding (cover, footer band and metadata only — the manual body is never altered), R2 storage and a draft for human review. If the source is not manufacturer / authorised, the job does not proceed.
Worker flow (Python, runs on the KVM 4 host):
source_url fetch OR direct upload
→ stream to disk → sha256 dedupe → extract
→ brand (cover / footer / meta) → upload R2
→ create draft → human review → publish

Optimised like it matters — because the host is modest

  • Bounded everything: 2 concurrent download workers, 1 PDF-processing worker, hard memory cap, streaming downloads to disk (never whole PDFs in RAM), resumable transfers — no dedicated nodes, no extra servers.
  • Cheap queries only: single indexed lookups by sha256 and (brand, model), prepared statements, batched inserts, no N+1, related links precomputed at publish — page renders never scan the catalogue.
  • Cache first: HTTP ETag / Last-Modified revalidation so re-checking a source URL you already imported costs a 304 instead of a full download, plus Cloudflare full-page cache + Redis object cache in front of WordPress.
  • Zero extra plugins: theme + HandbookNest Core + this worker script. That is the whole stack — nothing else to slow it down, update or conflict.
  • Polite and provenance-first: the worker only ever contacts the source URL you submitted — the manufacturer or authorised host — with per-host throttling, robots-respecting rates and a real HandbookNest user-agent with contact details. Every job logs its source URL and rights basis.

Design — the same, on purpose

Yes, the live site's design is this design — the preview is the design reference for the theme build, not a placeholder. It is deliberately not a generic template look: no stock hero-and-three-cards layout, no purple gradients, no glassmorphism. Instead it keeps what makes a manuals library work — search first, plain text, fast pages, nothing decorative between the reader and the manual — and dresses it in HandbookNest's own handcrafted identity: Fraunces serif headings with Inter body text, the warm bark / amber / clay palette, the nest-and-handbook mark, real contents lists with page numbers, and spec grids that read like a printed manual. Same layout, same components, same feel on WordPress — just wired to the Core plugin and the real catalogue.

SEO ground rules

Index control — the Indexing Control Center

  • The HTML handbook page is canonical; the PDF is a download, never a competing duplicate.
  • Sharded sitemaps split by brand and content type, submitted and status-checked through the Search Console Sitemaps API.
  • Index status syncs back via the URL Inspection API into an admin worklist; Search Analytics metrics sit alongside it. The Google Indexing API is not used for manual pages — it is documented for job postings and livestreams only.
  • Search results and thin brand × category combinations are noindex by rule.

Performance & search

  • Cloudflare in front, full-page cache, Redis object cache on the host.
  • Dedicated search (Typesense-class) — native WordPress search won’t survive model numbers and error codes at scale.
  • Bulk imports, OCR and PDF branding run on the pipeline, never on the web server’s request path.

Delivery timeline

Sat 14 Nov 2026

This preview. Branding applied, homepage + brand hub + handbook page designs, working catalogue search, the full blueprint — ready for your review.

In review
Next

Core plugin + database. Post types, taxonomies, custom tables and rewrite rules built and tested against the seed catalogue.

Queued
Then

Theme build. The templates from this preview, productionised in WordPress with the Fraunces / Inter type system.

Queued
Then

PDF pipeline + R2. Ingestion, OCR, branding and storage wired up; first real manufacturer PDFs processed end to end.

Queued
Launch

Seed import & go-live on handbooknest.com once the domain is registered, with sitemaps and search index in place from day one.

Queued