Cardstock today identifies a card from a photo and tracks cost and P&L by hand. The next release connects every card to live market data — automatic values at scan-in, a portfolio that reprices itself nightly, sell signals, and accounts that follow the user to any device. A subscription product, on the App Store.
We now hold API access to a professional card-pricing service covering sales data across the hobby — fair market values with confidence scoring, real comparable sales, price history at every grade, and market movement. Every feature on this board is built on capabilities we have already verified against our production key. It converts Cardstock's one manual step — "type in what you think it's worth" — into the product's strongest feature.
Scan it, know its value, track the position, act on the signal — from any device. Each pillar feeds the next.
AI vision reads the card — front and back — and fills every field.
A live fair market value attaches to the card the moment it's scanned.
The whole portfolio reprices itself nightly — real unrealized P&L, not guesses.
The app watches the market and tells the user when a card is worth selling.
Accounts in the cloud — the collection follows the user to every device.
Today the user scans a card, then goes off to research what it's worth. In the next release the review sheet answers that before they've typed anything.
The moment a card is identified, its fair market value — with a realistic range — is pre-filled into the record. Still editable, never required homework.
Raw, PSA 9, and PSA 10 values side by side at scan time. A raw card instantly shows its grading upside — before the user decides to keep, sell, or submit.
For graded cards, the cert number on the slab resolves the exact card, verifies the cert is genuine, and prices it at its exact grade — one step.
Actual recent sold prices — with dates, sources, and links to the sold listings — right in the review sheet. Today's eBay search link stays as the "see more."
When the match or the data is weak, the app says so and keeps the manual field front and center. Strong matches auto-fill quietly.
The portfolio is the screen users open every day and the one that sells the subscription — it deserves a real design decision, not a default. Five distinct directions, all powered by the same data layer: live values repriced nightly, sell signals, momentum, and liquidity. Scroll through and pick a lane.
Scroll sideways →
Misiorowski PSA 10 is in a selling window — up 22% this month on 9 sales this week. It would move fast at $464.Review →
Carter /99 crossed 2.6× your cost. You paid $120 in March; it books at $312 tonight.Review →
Rhodes Select is cooling — down 15% over 30 days on thinning volume. Worth deciding before it slides further.Review →
Ellison Optic RC is a grading candidate. Raw books at $38; a PSA 10 books at $210 against a ~$25 fee.Compare →
Directions 02, 03, and 05 aren't competing layouts — they're three altitudes of the same portfolio. The story of what happened, the numbers behind it, and the cards themselves. The unified screen stacks them in exactly that order down a single scroll: the user descends from "what happened" to "how it's going" to "what I own" with no tabs, no modes, no settings choice.
Misiorowski PSA 10 is in a selling window — up 22% this month on 9 sales this week.Review →
Carter /99 crossed 2.6× your cost. Paid $120 in March; books at $312 tonight.Review →
Hierarchy is scroll position, not navigation: story → numbers → cards. Each act peeks into the next so the screen always promises more below — no tabs, no mode switch to learn.
Every Brief insight is anchored to a real card. Tap it and the screen glides down to that card on the Shelf; flagged cards wear an amber dot. Same object, three altitudes — that's what makes three layouts feel like one room.
Once the Brief is read, it folds into a one-line title bar and the numbers lead. The Signals tile swells when signals exist and vanishes when there are none. Quiet days make a quiet screen.
Shared radii, one chip language, and the same card-art treatment across all three acts. The Brief's thumbnails, the Command Center's mover row, and the Shelf's grid are the same component at three sizes.
Nightly repricing writes a fresh Brief, so every morning opens with news instead of a form; pull-to-refresh reprices on demand. And the other two directions still ship inside this one: the Ticker's chart lives in the Command Center's value tile, and the Ledger becomes the "List" toggle on the Shelf. All five directions survive — four of them folded into a single screen.
One long scroll reads well but navigates poorly — "show me my cards" shouldn't mean a journey past two other sections. Four navigation models, each keeping the three-altitude idea while making every act one deliberate gesture away.
Scroll sideways →
Misiorowski PSA 10 — selling window open.
Decision: Option A — The Elevator, with stops named Brief · P&L · Shelf. One tap to any altitude on top of the approved composite, shipping in the flagship build (roadmap Phase 2). The Sheet Stack remains the documented evolution path if the screen later earns the deeper investment; the Pass Stack and Pages stay here as considered alternatives.
The detail screen stops being a static record and becomes a live market page for that exact card at that exact grade.
Hot, Rising, Stable, Cooling, or Falling — computed from real price movement and sales volume, shown as a badge with the percentage and timeframe.
A price sparkline for this card at this grade, in the same chart language as the P&L dashboard.
Every grade's current value with the user's own grade highlighted — the raw-vs-graded spread at a glance.
The last real transactions with dates, marketplaces, and links to the actual sold listings.
Every value states its evidence: confidence grade, sales count, data age. Weak data says so.
The player's whole market — sales volume and average price trend — as context under the card.
A new Market tab makes Cardstock a daily habit even on days the user isn't scanning — and the scanner itself gets a second brain.
The week's biggest price gainers across the hobby, filterable by sport and category, refreshed daily.
Movers the user holds are flagged in the feed. Market news becomes personal news.
Every scan is confirmed by two independent systems — our AI vision plus the market service's visual match. Agreement means near-certain IDs; disagreement gets flagged instead of guessed.
Scanned cards snap to a canonical catalog entry — consistent set names, variants, and numbers instead of freehand text.
Sign in with Apple, and the collection lives in the cloud — new phone, iPad, or the web, it's all there. Data is isolated per user at the database level, backed up daily, and built on the same infrastructure trusted by thousands of production apps.
Tiers and pricing get decided after the feature set locks, so packaging follows the product instead of constraining it. The economics are already validated and forgiving: variable cost per active user is cents on the dollar, so margins clear 90% at any realistic subscription price. Build first, package second.
Feature depth first, publication polish last. Phases 1 and 2 are the build — live values, then the full signal and market layer — all shippable to the current testers with no App Store review. Accounts and the store are the final step before the public launch.
Every feature on this board was checked against the pricing service's live API using our production key on July 2, 2026 — including batch valuation of 100 cards per request on our current plan, no upgrade required. Keys stay server-side in our existing secure gateway, responses are cached, and projected usage for the first cohort of users lands in the low hundreds of requests per day — comfortably inside plan limits. The one open item is deliberate: Phase 1 starts by validating automatic card-matching against our real inventory, so every later feature stands on measured accuracy, not optimism.