The quiet infrastructure play hiding in a to-do app
Cross-border sellers don’t lose money because they forget to reorder inventory. They lose it because the reminder lives in one tool, the SKU lives in another, and the person who was supposed to act on it is six time zones away with a different calendar. Every operator I know is running some Frankenstein stack of Slack pings, Notion databases, and half-abandoned Asana boards, and the actual “do this today” layer is usually whatever’s still open in a browser tab. So when an indie developer ships a task app whose entire pitch is local-first, no backend, no account, your data stays in your own iCloud, my ears prick up — not because I need another to-do list, but because the architecture underneath it is exactly the architecture more cross-border ops tooling should have. Cinch - Reminders & Tasks is a small product with a disproportionately interesting set of design decisions for anyone running a distributed commerce operation.
What Cinch actually is, and the problem it’s aiming at
Cinch is built by indie developer Petr Selinger, and his framing on the launch page is refreshingly honest: he liked the privacy and reliability of Apple Reminders but wanted “a simpler, less cluttered workspace for managing my day.” That’s the whole thesis. It is not trying to replace Todoist, Things, or TickTick. It is trying to be a better front-end for a backend you already own.
The specifics matter here, because they’re the entire product:
- No external backends, analytics, or custom servers.
- No account required — it connects directly to the Apple Reminders app you already use.
- All data stays strictly within your personal iCloud account.
- Built with SwiftUI, roughly 15 MB in size.
- Runs across iPhone, iPad, Mac (with keyboard-first support), Apple Watch, and Vision Pro.
- Includes Apple Intelligence features, extra-large widgets, and Shortcuts integration.
- A CLI for your agent: LLMs get direct access to your Reminders while data stays local.
- Free to use; a Pro tier is a yearly subscription covering one license for the whole family’s devices.
- Launch promo: code
PRODUCTHUNT50for 50% off the first year, redeemable through the App Store offer link.
Pricing for the Pro tier is, notably, not disclosed on the page beyond “a reasonable price.” That’s a small thing, but for operators who evaluate tools by seat cost across a team, it’s the first friction point.
Why this architecture is the actual story
Strip away the productivity-app framing and what Selinger has built is a local-first client with an agent-accessible CLI. That combination is the part cross-border sellers should be staring at. The reason most ops tooling fails inside a 5-to-15 person e-commerce company isn’t features — it’s that every tool wants to be the system of record, and nobody wants to migrate the system of record twice a year. Cinch sidesteps that by refusing to be a system of record at all. It’s a view layer on top of a data store Apple already syncs for you.
For a seller running a lean team across Shenzhen, Warsaw, and Austin, that’s not a minor convenience. It means the “who does what by when” layer doesn’t require onboarding, doesn’t require a per-seat contract, and doesn’t create yet another place where PII, supplier contacts, or payment-adjacent notes can leak. It also means the failure mode is graceful: if the app disappears tomorrow, your reminders are still in Apple Reminders. Compare that to the graveyard of ops SaaS that shut down and took everyone’s SOPs with them.
How it stacks up against what sellers actually use today
Let me be blunt about the competitive set, because “another task app” is a crowded category and the honest comparison is not against other to-do apps — it’s against the tools cross-border teams already pay for.
Against Notion: Notion is where most DTC brands keep their SOPs, supplier lists, and launch calendars. It’s a genuinely great database, and it’s genuinely terrible at nagging. Nobody opens Notion to be told what to do today. Cinch isn’t competing with Notion’s database; it’s competing with the Slack message that says “hey, did you file the Amazon Seller Central reimbursement claim yet?” That’s a real gap.
Against Todoist and TickTick: Both are more feature-complete for team workflows, both have proper cross-platform support including Android and web, and both have natural-language parsing that’s genuinely good. Cinch is Apple-only. If your ops lead uses a Windows laptop — and plenty do — Cinch is a non-starter. Selinger doesn’t pretend otherwise.
Against Apple Reminders itself: This is the honest comparison, and it’s where Cinch has to justify its existence. Reminders is free, ubiquitous, and already syncs. Cinch’s bet is that Reminders’ UI is cluttered enough that people will pay for a cleaner skin plus a CLI. That’s a real bet, but it’s a thinner moat than “we have a backend that does X.”
Against Motion and AI schedulers: Motion and its peers auto-schedule your tasks onto your calendar. That’s a fundamentally different value proposition — it’s not “clean view of my reminders,” it’s “an algorithm decides when I do things.” For a seller juggling supplier calls, ad spend reviews, and returns triage, auto-scheduling is often more useful than a prettier list. Cinch’s counter-argument is privacy: Motion needs your calendar and your task data on its servers. Cinch doesn’t.
Why Amazon sellers should care more than Shopify ones
Here’s a distinction I don’t see made often enough. Shopify operators tend to run their business out of a small number of dashboards — Shopify admin, Klaviyo, maybe Triple Whale. The work is inside those tools. A task app is nice-to-have.
Amazon sellers run their business across tools, because Amazon’s own surfaces are fragmented. You’ve got Seller Central for inventory and cases, Amazon Advertising console for spend, Helium 10 or Jungle Scout for research, a third-party FBA reimbursement tool, a repricer, and a returns dashboard. The actual job is remembering to check each one on the right cadence. That is precisely the workload a local-first, agent-accessible reminder layer is built for. If you’re an Amazon operator, the CLI-to-your-LLM feature is the one worth testing this week — not the widgets.
What cross-border sellers can actually borrow from this
Even if you never install Cinch, the design choices are worth stealing for your own internal tooling and vendor evaluation.
1. Refuse to own the data. The single most underrated line in the launch copy is “no external backends, analytics or custom servers.” Every SaaS vendor you evaluate this year should be asked: where does my data live, who can see it, and what happens to it if you get acquired or shut down? For sellers handling supplier contracts, MOQs, and margin data, that question is not paranoia — it’s basic operational hygiene. Tools that store your data in your cloud account (iCloud, Google Drive, your own S3 bucket) are structurally safer than tools that store it in theirs.
2. Give your agents a CLI, not an API key. The “CLI for your Agent” feature is the most forward-looking thing on the page. The pattern — expose your internal data to an LLM through a local command-line interface rather than a cloud API — is going to become standard for privacy-sensitive ops. If you’re building internal tooling with Claude or ChatGPT, consider whether the model needs to reach your data over the network at all, or whether it can run locally against a CLI. The latter is slower to set up and dramatically easier to defend in a security review.
3. One license, all devices, family sharing. The family-license model is a small detail with real implications for small teams. Most B2B SaaS prices per seat, which punishes exactly the behavior you want — everyone on the team being able to see and act on the same tasks. A family license is a hack around per-seat pricing that works for teams up to about six people. If you’re a 3-5 person DTC brand, this is a meaningful cost difference versus a $12/seat/month tool.
4. Cross-device coverage is table stakes, not a feature. Cinch’s device list — iPhone, iPad, Mac, Watch, Vision Pro — looks impressive until you notice what’s missing: Android, Windows, and web. For a globally distributed team, that’s a real limitation. The lesson isn’t “Cinch is bad,” it’s “audit your own stack for the same gap.” If your ops tool doesn’t work on the device your warehouse lead actually carries, you don’t have an ops tool.
Where the math breaks
Let me do the unglamorous arithmetic on the “family license” pitch. If your team is six people and a competitor charges $10/seat/month, that’s $720/year. If Cinch Pro is, say, $40-60/year for the family (pricing not disclosed, so I’m guessing from the “reasonable price” framing), you’re saving $650+. That’s real money for a bootstrapped brand.
But the math flips hard the moment you need one thing Cinch doesn’t do: shared task assignment with permissions. Apple Reminders’ sharing model is built for families, not for teams with role-based access. If you need “only the finance lead can close the reconciliation task,” you’re back to a proper B2B tool. The family-license savings only hold if your team is small and trust-based, which describes a lot of early-stage DTC brands and almost no mature ones.
Where my judgment says it falls short
I like the architecture. I’m skeptical about the product-market fit for cross-border sellers specifically, for three reasons.
First, the Apple-only constraint is disqualifying for most cross-border teams. Cross-border means, by definition, people in different countries on different devices. Your sourcing agent in Guangzhou is on Windows or Android. Your 3PL contact in Rotterdam uses whatever their employer issued. An Apple-only task layer can be the founder’s personal system, but it cannot be the team’s system. Selinger is upfront that this is an Apple-ecosystem play, and that’s fine — but it caps the addressable market at “Apple-using solo operators and very small teams,” which is not where most cross-border complexity lives.
Second, the agent CLI is underspecified. “Give your LLMs direct access to your Reminders while keeping all data local and private” is a great sentence and a vague product. Which LLMs? Local models only, or can Claude Desktop reach it? What’s the permission model? Can a prompt-injected agent delete reminders? The launch page doesn’t say, and for a feature positioned as a differentiator, that’s a gap. I’d want to see documentation before trusting it with anything operationally important.
Third, the moat is thin. A cleaner UI on top of Apple Reminders is a weekend project for a competent SwiftUI developer. The CLI is more defensible, but Apple could ship a Reminders CLI tomorrow and the entire value proposition evaporates. Selinger’s protection is speed and taste, not technology. That’s a legitimate indie strategy — it’s how a lot of good small software gets built — but it means buyers should treat Cinch as a tool, not a platform. Don’t build your ops process on top of it.
The Vision Pro detail is a tell
I want to flag one thing that most coverage will skip: Cinch runs on Vision Pro. That’s a deliberately maximalist device list, and it tells you who Selinger is building for. This is a product for Apple power users who want their entire ecosystem to feel coherent. It is not a product designed around the messy, cross-platform, low-trust reality of international commerce ops. That’s not a criticism of the product — it’s a clarification of who should and shouldn’t care.
What I’d watch / test next
If you’re a cross-border operator reading this, here’s what I’d actually do this week.
Test the CLI against a real workflow. Install Cinch, create a Reminders list called “Ops — Daily,” and populate it with the five recurring checks your business actually needs (FBA reimbursement window, ad spend vs. ACOS threshold, restock trigger, returns aging, supplier payment due). Then wire it to a local LLM and see whether the agent can read and act on it without leaking data. That’s a two-hour experiment with a potentially large payoff in how you think about agent access to ops data.
Audit your own stack for the “where does the data live” question. Make a list of every tool that touches supplier contacts, margin data, or customer PII. For each, write down whether the data lives in your account or the vendor’s. You will almost certainly find two or three tools you should replace with self-hosted or local-first alternatives, regardless of what you decide about Cinch.
Use the promo code if you’re Apple-only. PRODUCTHUNT50 gets you 50% off the first year via the App Store redemption link. At that discount, the “is it worth it” question gets much easier for a solo operator or a two-person brand. Just don’t let it become your team’s system of record — that’s not what it’s built for, and Selinger isn’t pretending otherwise.
The bigger takeaway isn’t about Cinch. It’s that the local-first, agent-accessible, no-backend pattern is coming for ops tooling, and the cross-border sellers who internalize that pattern early — even if they never install a single app that uses it — will be the ones whose stacks survive the next round of SaaS consolidation.
