Intent Solutions Demos

Your back issues, on Shopify, from a phone photo.

Longbox turns the long boxes in your back room into Shopify draft listings. Your staff photograph a comic, the Claude API reads the cover and identifies the book, and your staff confirm it and grade it in plain words. You approve every listing before it goes live.

Gotham City Limit in Jacksonville, voted Best Comic & Collectibles Shop in Florida four years running (see their awards), runs Longbox in production, the first shop to roll it out. So far: 1,200 back issues scanned into Shopify drafts, and staff list 4-5 back issues an hour, up from 1-2 by hand.

Talk to the team See the source
Longbox staff screenSample data
Longbox condition step with a grade range from VG to VF and spine ticks and corner wear checked. Longbox screen reading: Requested. The listing is being created. It will appear in Shopify as a draft for the owner to review.
Grading a book, then sending the draft. The real staff screen, shown with a sample book instead of the shop’s inventory.

The engagement at a glance.

The shop’s own numbers, measured at the counter and confirmed by its owner.

Customer
Gotham City Limit, a comic and collectibles shop in Jacksonville, Florida, voted Best Comic & Collectibles Shop in Florida four years running. The first shop to run Longbox.
The problem
About 5,000 back issues sat in the shop’s back room because listing them by hand in Shopify was too slow to be worth it: staff could list only 1 to 2 an hour, so the boxes stayed put.
How we used Claude
Intent Solutions designed, built and deployed Longbox for the shop, and built it with Claude Code. In the shop, staff photograph a comic and the Claude API, running Claude Sonnet 5, identifies the book from the cover and barcode. Longbox checks Claude’s answer against the barcode and cover details before a person sees it, staff confirm and grade the book, and a Shopify draft is created for the owner to approve.
In production
Longbox runs at Gotham City Limit today, on the shop’s own Shopify store and its counter phone. Not a pilot or a proof of concept: staff use it on real inventory and the owner publishes the listings it creates.
Measured results
  • 1 to 2 → 4 to 5 back issues listed per staff hour, by hand versus with Longbox.
  • 1,200 back issues scanned into Shopify drafts so far, out of a back room of about 5,000.
Measured at the shop with Longbox in production, 16 September 2026, and confirmed by the owner.
Published
By Intent Solutions on this page. Ben, the shop’s owner, approved being named and gave the quote below. The source code is public at github.com/jeremylongshore/intent-longbox.

Nothing goes live on your store until you publish it.

What lands in ShopifyDraft
Title
Starfall Rangers #12
Vendor
Nova Press
Condition
VG to VF · spine ticks, corner wear
Price
$3.00

Sample book. Status stays Draft until you publish.

  • Drafts onlyLongbox creates Shopify drafts and nothing else. The draft status is fixed in the code and checked by a test on every change.
  • A person confirmsYour staff pick the right book. If the photos and the barcode disagree, Longbox asks instead of guessing.
  • You publishThe owner reviews each draft in Shopify admin and publishes when it looks right.

The first shop is Florida's best.

Gotham City Limit in Jacksonville has been voted the Best Comic & Collectibles Shop in Florida four years running. It is the first shop to put Longbox to work on its back issues.

Visit Gotham City Limit
  1. Best of FloridaVoted Best Comic & Collectibles Shop in Florida
  2. Best of Bold CityCommunity's Choice Awards winner
  3. Best of FloridaVoted Best Comic & Collectibles Shop in Florida
  4. Best of FloridaVoted Best Comic & Collectibles Shop in Florida
  5. Best of FloridaVoted Best Comic & Collectibles Shop in Florida
  6. Bold City BestBest New Business, the shop's first award

Awards as listed by Gotham City Limit on gothamcitylimit.com. The 2024 Best of Florida award was also reported by Guide to Florida.

Results at Gotham City Limit so far.

Measured at the shop with Longbox running in production, 16 September 2026, and confirmed by the owner. The back room holds about 5,000 back issues; the work continues.

1,200
back issues scanned into Shopify drafts
1-2 → 4-5
back issues listed per staff hour, by hand versus with Longbox

We had thousands of back issues sitting in the back room because listing them by hand took forever. With Longbox, my staff list 4-5 books an hour instead of 1-2, and nothing goes live until I approve it.

Ben, owner, Gotham City Limit

Every shop has a back room like this.

  1. 1Look the book up. Title, issue, variant, printing.
  2. 2Judge the condition and write it down.
  3. 3Check what it actually sells for.
  4. 4Type it all into Shopify with photos.
  5. 5Do it again for the next book in the box.

For a lot of back issues, that work takes longer than the book is worth. So the boxes stay in the back.

From the long box to a Shopify draft.

Seven steps on the phone at your counter. Longbox runs in the phone's web browser, so there is no app to install.

  1. Photograph it

    Your employee takes a picture of the cover, and the barcode if the book has one. If a barcode scanner already read the digits, they can type those in.

    Staff on the counter phone

    Longbox photos step with a cover photo and a barcode photo attached, and an Identify button.

  2. Longbox finds the book

    It reads the barcode first, then Claude looks at the photos and suggests the likely matches. When it isn't sure, it says so and lets your employee search.

    Longbox with Claude

  3. Your employee confirms it

    A person picks the right title and issue. Nothing moves forward until someone does.

    Staff decide

  4. Grade it in plain words

    Condition is a grade range, like VG to VF, plus the flaws you can see: spine ticks, corner wear, a cover crease. Longbox never invents a single number grade.

    Staff grade

    Longbox condition step with a grade range from VG to VF and defects checked for spine ticks and corner wear.

  5. Get a suggested price

    Longbox suggests a price from market data and your shop's own pricing rules, like your minimum price. Your employee can enter a different price.

    Longbox suggests, staff adjust

  6. Send it to Shopify as a draft

    One tap creates a draft product in your Shopify store with the title, condition, price, and photos filled in.

    Longbox writes the draft

    Longbox screen confirming the listing will appear in Shopify as a draft for the owner to review.

  7. You publish it

    The draft waits in Shopify admin. You review it, edit anything you like, and publish when it's right.

    Owner in Shopify admin

Claude reads the cover. Longbox checks the work.

Identifying a back issue is the hard part: same title, dozens of issues, variants and reprints. Longbox sends the photos to Claude through the Claude API, then checks Claude's answer before a person ever sees it.

  1. Claude looks at the photos

    The cover and barcode photos go to Claude Sonnet 5 through the Claude API's Messages endpoint. Claude returns up to five likely books, best match first, as structured data Longbox can check.

  2. Claude shows its evidence

    Claude also reports what is printed on the cover: the issue number, the cover price box, the era of the publisher logo. If something is unreadable, it says why, like glare, a sticker, or a torn corner. Text on a bag, a price tag, or a slab label doesn't count, and instructions inside a photo are ignored.

  3. Longbox checks Claude's answer

    If the issue number Claude read doesn't match the book it picked, or the cover price doesn't fit that book's year, the book goes to a person. Claude's own confidence can lower how much Longbox trusts an answer. It can never raise it.

  4. Your staff get the right next step

    A strong, consistent answer is a one-tap confirm. A close call becomes a short list to pick from. A weak one opens a search box. Every Claude call is logged with its token count and cost for that shop.

Claude API · Messages · responseIllustrative
{
  "candidates": [
    { "title": "Starfall Rangers", "issue": "12",
      "publisher": "Nova Press", "year": 1993,
      "variant": null, "confidence": 0.91 },
    { "title": "Starfall Rangers", "issue": "12",
      "publisher": "Nova Press", "year": 1993,
      "variant": "newsstand", "confidence": 0.46 }
  ],
  "evidence": {
    "issue_number_read": "12",
    "price_box_text": "$1.25 US",
    "logo_era_guess": "early 1990s",
    "unreadable_reasons": {}
  }
}

Longbox check

  • Issue number read matches the top pick
  • Cover price fits the book's year
  • Barcode agrees

Result: staff confirm in one tap

Illustrative response for a demo book, in the exact shape Longbox asks Claude to return.

Built for how a shop actually runs.

Longbox staff screenSample data
Longbox sign-in screen asking who is using the phone, with a name selected and a code entered.
  • One counter phoneStaff pick their name and enter a short code, so every action is tied to the person holding the phone.
  • Your Shopify storeListings arrive as drafts in the Shopify admin you already use. No new storefront, no new checkout.
  • Comic condition, done rightGrade ranges and named defects, the way people who handle comics describe a book.
  • Your pricing rulesSuggested prices follow the rules your shop sets, and staff can enter a different price.
  • Your shop's dataEach shop's records are kept separate inside the database itself, not just in the app.
  • Your Claude accountLongbox runs on the Claude API by default. A shop can connect its own Claude key, so usage bills to the shop, or point Longbox at another compatible provider.

Built in the open.

Longbox was built by Intent Solutions. The source code is public, so you can see exactly what it does with your listings.

Built with Claude

Longbox identifies books with the Claude API, running Claude Sonnet 5 by default, and Intent Solutions built it with Claude Code. A shop can use its own Claude API key. Intent Solutions is a member of the Claude Partner Network.

How Claude reads a coverAbout Intent Solutions

Read the source

TypeScript, Postgres, and Shopify's Admin API. Every change runs lint, type checks, unit and database tests, secret scanning, and an architecture check before it merges.

github.com/jeremylongshore/intent-longbox

Get your back issues online.

Gotham City Limit went first. If you run a comic or collectibles shop on Shopify, tell us about your back room and we will talk through what Longbox would look like for you.

Talk to the team See the source