browser · frame · project

Vibe Kanban

A local web kanban where issues become agent workspaces with diffs, previews and PRs; its company shut down in April 2026 and the project is community-maintained.

inspireevidence: strongby BloopAI (bloop)www.vibekanban.com ↗

Vibe Kanban

  • Maker: BloopAI (bloop)
  • URL: https://www.vibekanban.com, https://github.com/BloopAI/vibe-kanban
  • Status (10 October 2026): sunsetting. bloop announced it was shutting down as a company on 10 April 2026; the project continues as open source, community maintained. Rust + TypeScript, Apache-2.0, 28,315 stars, repository created 14 June 2025, last push 19 September 2026. Not in our first pass; added because its shutdown is evidence about the business of agent frames.

What it is

A local web app (npx vibe-kanban) where you plan work as kanban issues, then start a coding agent (Claude Code, Codex, Gemini CLI, Copilot, Amp, Cursor, OpenCode and others) in a workspace with a branch, terminal and dev server; review diffs with inline comments sent back to the agent; preview the app in a built-in browser; open and merge PRs.

The problem it’s solving

“Software engineers spend most of their time planning and reviewing coding agents,” so make planning and review faster.

Its path / bet

Put the task board at the centre and treat agents as workers you assign issues to. A browser UI served locally, so no desktop install.

How it works

Issues on a board; each started issue becomes a workspace (worktree, terminal, dev server) running an agent; review happens in the same UI, with comments fed back to the agent.

Strengths

  • Very fast adoption (28,000 stars in under a year) and daily use by thousands, per the makers.
  • The board-first framing makes the person a planner and reviewer, which is the real job.

Weaknesses / limits

  • The company “couldn’t find a business model that we could get excited about”; most users were free. Future maintenance depends on the community.

Relation to fictty

Inspire (and caution). Two lessons. The frame layer is crowded and hard to monetise: if a 28,000-star frame couldn’t sustain a company, a content runtime won’t by being a product either. And the board is a screen an agent could put up and keep live from data; it doesn’t need to be an app.

Could fictty adopt it instead of building?

No; it’s a web app frame, now community-maintained.

What fictty should take from it

  • Keep fictty a small open runtime with no service to sustain; integrate with frames rather than becoming one.
  • A kanban view bound to gh issue list --json or a task CLI is a good fictty demo of “an app’s main screen as one value”.

Sources