AI-native planning · Developers only

Stop writing
specs.
Start shipping.

Describe any task in plain English. KanPlan turns it into a full implementation plan — acceptance criteria, scope, technical notes — in seconds. Smart tags steer every prompt.

50 free AI generations/mo · No credit card required

my-saas-appBoard
Ideas3

OAuth with GitHub

auth · backend

Dark mode

frontend

+ Add card

Tasks2
AI generated

Rate limit middleware

security

Settings page

frontend

Done2
Shipped

CI/CD pipeline

Shipped

Auth setup

AI Generated · 2.4 s

Full spec · Criteria · Edge cases

The old way vs. the KanPlan way

Before KanPlan

  • 30+ min writing out task specs from scratch
  • Forget to add acceptance criteria until QA flags it
  • New project setup takes half a day
  • Team writes cards inconsistently — different levels of detail
  • Jump between Notion, Jira, docs, and chat to get context
  • Start a new sprint and rebuild the same board structure again

With KanPlan

  • Type one sentence — AI generates the full spec in ~3 seconds
  • Acceptance criteria, scope & technical notes always included
  • Apply a blueprint — board is ready in one click
  • Tags ensure every AI card follows your team's standards
  • All context lives on the card — no context switching
  • Export your board config as a blueprint and reuse forever

01 — Import Your Existing Work

Already have a project?
Bring it with you.

Switching tools is only painful when you have to leave everything behind. KanPlan lets you import your existing GitHub Issues — titles, descriptions, labels, milestones — directly into a KanPlan board, with column mapping you control. No copy-pasting. No rebuilding from scratch.

Once imported, every card becomes a first-class KanPlan card. You can immediately run AI generation on any of them, apply your tag library, and slot them into a blueprint workflow — all without losing a single issue.

GitHub Issues importConnect your repo, select open issues, map them to columns. Done in under 60 seconds.

Labels → TagsYour GitHub labels are auto-mapped to KanPlan tags with matching colors.

Milestones → ColumnsMilestones can optionally map to board columns — or you choose the mapping yourself.

More sources comingLinear, Jira, Notion, CSV and more planned. Import once, stay in KanPlan.

No lock-in — all your data exports as JSON anytime.

github.com / acme / my-saasConnected

Open Issues — select to import

#142Add OAuth with GitHub
authbackend
#139Rate limiting on /api/*
securitybackend
#137Settings page — profile tab
frontend
#134Email notifications — Resend
backendnotifications
#131Dashboard analytics charts
frontenddata

Column mapping

open + no milestoneIdeas
open + milestone: v1.0Tasks
open + label: in-reviewIn PR
closedDone
Import 3 issues Edit mapping

Coming soon

LinearJiraNotionTrelloCSV / ExcelAzure DevOps

02 — AI Card Generation

One sentence in.
A full plan out.

You describe the task. KanPlan's AI writes the implementation steps, acceptance criteria, technical scope, and edge cases — then shows a side-by-side diff so you can see exactly what it added before committing.

You're never locked into the output. Review the diff and choose:

AcceptSave as-is — version history created automatically

Edit inlineMake changes inside the diff before saving

RegenerateAdd a clarifying note and run AI again

RejectDiscard — your original text is untouched

Real examples — one sentence becomes:

Add rate limiting to the API

Token-bucket rate limiting on /api/* routes. 100 req/min per IP. Redis-backed with sliding window fallback. 429 + Retry-After header on breach. Tests: burst scenario, TTL expiry.

Build the user settings page

Settings page with tabs: Profile, Security, Notifications. PATCH /users/:id endpoint. Optimistic UI. Zod form validation. Unsaved-changes prompt on navigate-away.

Set up email notifications

Transactional email via Resend. Queue-backed with retry logic. Templates: welcome, password reset, billing alert. Unsubscribe link per GDPR. Integration test: mock SMTP delivery.

Live Diff Preview

title: "Rate Limiting Middleware"

column: "Tasks"

− description: "Add rate limiting to the API"

+ description: "Token-bucket limiting on /api/* routes"

+ implementation:

+ - Redis token bucket, 100 req/min per IP

+ - 429 with Retry-After header on breach

+ - Sliding-window fallback on Redis outage

+ acceptance: "Burst test passes; TTL expiry resets"

+ scope: "middleware/, redis.ts, api/* routes"

AcceptEditRegenerate

AI Providers

OpenAIAnthropicGoogleOllamaLM StudioCustom

Model scope

Set a project default, override per column or per card — mix providers freely.

Developers report spending 20–40% of planning time writing specs and acceptance criteria. KanPlan does that in seconds, so you spend your energy on the actual work.

02

Tag Library — prompt injection

React

Use React 19. Prefer Server Components for data-fetching. Use client components only for interactivity.

Security

Follow OWASP top 10. Validate all inputs. No raw SQL. Use parameterized queries.

TypeScript

Strict mode. No implicit any. Use Zod for runtime validation of external data.

Testing

Unit tests with Vitest. Integration tests for API routes. Target 80%+ coverage.

Accessibility

WCAG 2.1 AA. Semantic HTML. ARIA labels where needed. Keyboard navigable.

every active tag's prompt is appended to each AI generation

Tag stacking example

Card: “Add user avatar upload”
Tags active: React + Security + TypeScript

AI automatically: uses Server Actions for the upload, validates MIME type and file size per OWASP, types the handler with strict TypeScript, and adds sanitization notes.

No extra prompting needed — the tags did the work.

03 — Smart Tags

Your coding standards,
encoded once.

KanPlan tags are not labels for sorting. Each tag holds a prompt template — a plain-language instruction that gets automatically injected into every AI generation where that tag is active.

This means you define your team's standards once — “we use React 19”, “we always write Zod schemas”, “WCAG AA is non-negotiable” — and the AI applies them consistently to every card, forever.

Stack multiple tags and they compound: a card with “React + Security + Testing” produces a spec covering component architecture, security considerations, and test cases — without any extra prompting from you.

Icon + color-coded for instant visual identification

Import / export as JSON to share standards across teams

Assign per card or per project

Variables support — dynamic values per tag template

Free: 5 tags · Basic: 20 tags · Pro: unlimited

04 — Blueprints

Build your perfect board once.
Never rebuild it again.

How many times have you set up the same board structure for a new project? SaaS starter columns, the same tags, the same model — again and again. A KanPlan blueprint captures everything: column names, all tags with their prompt templates, default AI model, and per-column model overrides.

Apply a blueprint to any new project in one click. Choose Replace to overwrite entirely, or Merge to layer on top of what you have. Export as .blueprint.json to share with your team or check into version control.

Snapshot any project

Turn a live board into a reusable blueprint

One-click apply

Bootstrap new projects instantly

Replace or Merge

Overwrite or layer on existing config

Export .blueprint.json

Share, version-control, and publish

SaaS StarterGPT-4o

Full-stack web app from first idea to production deploy

Ideas → Specs → Tasks → Build → In PR → Ship

ReactTypeScriptSecurityTesting
Bug TriageClaude 3.5 Sonnet

Structured bug lifecycle from report to verified fix

Reported → Confirmed → Fixing → In PR → QA → Closed

DebuggingSecurityTesting
API ServiceGPT-4o

Design-first REST API build with full documentation

Design → Implement → Test → Document → Deploy → Monitor

API DesignSecurityTypeScriptTesting
Mobile AppClaude 3.5 Sonnet

Cross-platform mobile from concept to app store

Concept → Wireframe → Build → Test → Submit → Live

React NativeA11ySecurity

05 — The Kanban Board

A board that keeps
up with how you work.

Two views, every workflow covered. The Kanban view gives you a visual overview of your sprint — drag cards between columns with optimistic updates. The List view is a power-user table: sort, bulk-assign, collapse status groups, and filter down to exactly what you need.

6-column pipeline

Ideas → Specs → Tasks → In Progress → In PR → Done. Rename any column to match your process.

Drag & drop

Native drag with optimistic UI. Cards snap into place immediately; sync happens in the background.

Card versioning

Every AI accept and every manual save creates a snapshot. Revert any card to any previous state.

List view

Sortable table mode with bulk selection, collapsible status groups, and the full card editor.

Per-card context

Title, description, implementation steps, acceptance criteria, scope, history, and AI version log — all on one card.

Ideas
Auth flow
Search UI
Dark mode
Specs
API v2
Rate limit
Tasks
DB migration
Email queue
In Progress
OAuth
Uploads
In PR
Nav refactor
Done
CI pipeline
Logging
Auth tests
List ViewSort · Filter · Bulk select
OAuth 2.0 with GitHubIn Progress
Rate limiting middlewareTasks
Database schema migrationTasks
CI/CD pipeline setupDone

Real workflow example

Zero to sprint-ready in minutes

Here's how a developer kicks off a new feature project with KanPlan — from blank board to a fully-specced sprint backlog.

1

Apply the SaaS Starter blueprint

Open a new project, click Apply Blueprint → SaaS Starter. Instantly get 6 well-named columns, the React + TypeScript + Security + Testing tag library (with pre-written AI prompts), and GPT-4o set as the project default.

Board ready in one click. No configuration. Nothing to name or wire up.

2

Drop rough ideas into the Ideas column

Type quick one-liners: "User auth with GitHub OAuth", "Dashboard analytics charts", "Export data as CSV", "Email notification system". These are stubs — no structure needed yet.

Takes ~2 minutes. Capture ideas as fast as they come.

3

Generate implementation plans with AI

Click Generate on each card. The AI — guided by your React, Security, and TypeScript tags — expands each one-liner into a full spec: implementation steps, edge cases, acceptance criteria, affected files.

~3 seconds per card. All 4 cards fully specced in under a minute.

4

Review diffs, accept or refine

Side-by-side diff shows exactly what AI added. Accept the ones that look right. For "Email notifications" add an override — "Use Resend, not Sendgrid" — and regenerate. Accept the updated version.

You stay in control. AI is a co-author, not an autopilot.

5

Move cards to Specs and start the sprint

Drag clear cards to Specs. Leave ambiguous ones in Ideas for another pass. The team now has a consistently-written, fully-detailed backlog — with no Thursday-afternoon spec-writing sessions required.

Total time: ~5 minutes. Same output that used to take half a day.

Pricing

Start free. Scale when ready.

Free

$0

Forever free

  • 50 AI card generations / mo
  • 5 smart tags
  • Unlimited projects
  • All 6+ AI providers
  • Card versioning
Get started

Basic

$9.95

per month

  • 200 AI card generations / mo
  • 20 smart tags
  • Import / export blueprints
  • Snapshot projects as blueprints
  • Priority AI response
Start Basic

Pro

$29.94

per month

  • 1 000 AI card generations / mo
  • Unlimited smart tags
  • Usage-based billing
  • Team blueprint sharing
  • All future features
Start Pro

Your next project starts here.

Free to start. No credit card. Your first 50 AI generations are on us.