Should I vibe code
Run kanban boards with cards, checklists, labels, dates, and simple automation
Kanban is drag-and-drop over an ordered list. This one is genuinely within reach.
?
Their verdict, the Premium price and the build-time estimate come from their entry, MIT-licensed. Checked 2026-08-03.
?
Our verdict, the regret score and everything below it. Editorial and unsponsored — nobody can pay to be moved.
The honest answer
why the verdict is what it is
The single most buildable project management tool, because the data model is small. Real-time updates for a team are the only meaningful extra.
What actually breaks
not "if". the specific failures.
- Card ordering, stored as an integer position, so every drag rewrites half the column and two people dragging at once produce a shuffle
- Realtime, because a board that needs a refresh to show a colleague's change stops being a shared board
- Offline and flaky connections, where the optimistic move you showed the user silently did not happen
- Mobile drag and drop, which is a genuinely hard interaction and the one people use most
- Notifications and activity history, which are the entire reason anyone opens the board on Monday
The demo was beautiful within a day. Three weeks in, two people move cards during the same standup and the column ends up in an order neither of them chose — because position is an integer, both clients computed 'between 3 and 4', and the last write won twice. Nobody lost data. They just stopped trusting the board, which for a kanban tool is the same thing, and quietly went back to a group chat.
Is that you?
the verdict is a default, not a law
- It is a single-user board and you are the only one who moves a card
- It is a read-only view of work that lives somewhere else
- Reordering is a text field or a sort, not drag and drop
- More than one person moves cards on the same board
- Position is an integer or an array index
- There is no activity log to explain what changed overnight
- People are meant to use it on a phone
If you build it anyway
the checklist, then the prompt that enforces it
- Order with fractional ranks — a string or decimal key between neighbours — so a move writes one row and concurrent moves cannot collide destructively.
- Pick realtime or explicit refresh and be honest about it. A stale board that looks live is worse than one with a visible refresh button.
- Every mutation is an append-only activity event. The board state is a projection, and 'what changed' is the feature people actually return for.
- Reconcile optimistic updates against the server response and visibly revert on failure. Never leave a card where the user dropped it if the write failed.
- Build the mobile interaction early or declare it out of scope. Discovering drag and drop on touch at the end is how these projects stall.
Before you build a kanban board, apply these and push back if I ask you to break them. 1. Ask me whether more than one person will use the same board. If yes, tell me that ordering and realtime are the whole difficulty and that everything else is a weekend. 2. Do not store card position as an integer or array index. Use a fractional rank — a decimal or a lexicographic key computed between the neighbours — so a move updates exactly one row. Show me what happens when two clients move different cards to the same slot simultaneously. 3. Record every change as an append-only activity event with actor and timestamp, and derive board state from it. Build the activity feed in the first version; it is why people open the board. 4. If the UI updates optimistically, reconcile against the server response and visibly revert on failure. Never leave a card sitting where the user dropped it when the write did not land. 5. Choose explicitly between realtime sync and manual refresh, and tell me the cost of each. If we skip realtime, make staleness visible in the UI rather than pretending. 6. Decide about mobile now. Touch drag and drop is a real piece of work — either schedule it or put 'desktop only' in the README. 7. Out of scope unless I ask, and say so: attachments, comments, labels filtering, swimlanes, automation rules, notifications by email. 8. Tell me plainly which part of Trello I am not building, so I can decide before rather than after.
That one keeps you out of trouble. For the prompt that actually builds it, canivibecodeit.com has one.
their build prompt ↗Or don’t build it
the boring option, and the way back out
The moment a second person depends on the board. $12.50 a month buys realtime, mobile, notifications and an ordering algorithm that survives two people dragging at once — none of which show up in the demo you will build in an afternoon.
$12.5/mo is cheaper than your weekend.
Keep boards, cards and the activity log as plain exportable rows with stable ids, and keep the rank field human-readable. Trello and most competitors import CSV, so an export that includes list, position and timestamps means the team can move without retyping anything.
Active open-source project and issue tracking platform.
Questions
What is a fractional rank and why does it matter so much?
Instead of position 1, 2, 3 you store a value between the neighbours — 1.5, or a string like 'aM'. Moving a card writes one row instead of renumbering the column, and two people moving different cards cannot overwrite each other's ordering. It is a small change that removes most of the concurrency problem.
Is a single-user kanban board worth building?
Genuinely, yes — that is the first SHIP IT condition. Almost everything hard here comes from concurrency and mobile. Alone, on a desktop, a board is a list with a drag handle, and building one you actually like is a good weekend.
Every week, someone ships something they shouldn’t have.
New verdicts, the worst thing that landed in the trap, and the occasional incident report. No other email, ever.
Project management tools are easy to build and impossible to get a team to adopt.
Configurable boards for one team is achievable. Configurable boards for any team is a platform.
Tasks, docs, goals, dashboards and automations. That is five products and you want all five.
last reviewed 2026-08-03 · verdict is editorial and unsponsored · shared entry data from canivibecodeit under MIT · not legal advice