TL;DR

Your kanban board is full of ticket titles. Nobody remembers what "Refactor auth module" actually means two weeks later. Tekk.coach fixes that. Each card links to a full AI planning session that read your codebase before writing anything. You get context, scope, subtasks, file references — not just a sticky note. Start free, no setup required.

Try Tekk.coach Free →


How Tekk.coach Does Kanban

Most kanban board tools give you columns and cards. Tekk does too — To Do, In Progress, Done. Three columns, drag and drop. Simple on purpose.

The difference is what's inside each card. Before Tekk's AI writes a plan, it connects to your GitHub, GitLab, or Bitbucket repo via OAuth and reads your codebase. Semantic search, file search, regex, directory browsing, repo profiling — it builds an actual picture of your project before suggesting anything. Then it asks you clarifying questions. Then it writes a spec.

That spec streams into a rich text editor as a living document. Not a chat transcript. Not a locked artifact. An editable plan that includes a TL;DR, explicit scope (what's being built and what's not), subtasks with acceptance criteria tied to real files in your repo, assumptions with risk levels, and validation scenarios. This is what lives inside a Tekk kanban card.

When you reopen a card three weeks later, you don't have to reconstruct your thinking from a ticket title. The plan is there. When you hand a task to an AI coding agent — Cursor, Claude Code, Codex — the context is there too.


Key Benefits

Context that doesn't disappear. Every card holds the full reasoning behind a task. No more reconstructing decisions from memory or Slack threads.

Codebase-aware specs. The AI doesn't plan in a vacuum. It reads your repo, knows your file structure, and writes subtasks that reference actual files. Generic advice is not an option.

No ceremony. No sprint planning, no backlog grooming, no custom field configuration, no approval chains. Open a task, get a plan.

Works with your AI coding tools. If you're building with Cursor, Claude Code, Codex, or Gemini, Tekk is the planning layer that gives those tools real context before they touch your code.

Free to start. No credit card, no seat minimums. Build your first spec in minutes.


How It Works

  1. Connect your repo. Link GitHub, GitLab, or Bitbucket via OAuth. Tekk indexes your codebase immediately.

  2. Create a card. Add a task to your kanban board — any technical intention works. "Add rate limiting to the API." "Migrate from REST to tRPC." Whatever you're building next.

  3. Run the AI planning session. Tekk's multi-turn workflow: searches your codebase, asks clarifying questions, presents options, then writes the full spec.

  4. Get a complete plan. TL;DR, build/no-build scope, subtasks with acceptance criteria and file references, assumptions with risk ratings, validation scenarios.

  5. Move the card. Work through your kanban board with actual specs behind every column. Or pass the plan to your AI coding agent and let it execute with real context.


Who This Is For

Solo founders and indie builders. You're the architect, the PM, and the developer. You don't have a team to rubber-duck with. Tekk is the thinking partner that knows your codebase before it opens its mouth.

Small teams building with AI coding agents. If your stack includes Cursor, Claude Code, or Codex, you know the problem: the agent writes code, but the planning context lives in your head. Tekk closes that loop. Each card is a brief your AI agent can actually use.

Developers who are done with Jira. You don't need sprint velocity tracking, custom workflows, or approval chains. You need to know what to build, why it matters, and what files to touch. Tekk gives you that without the overhead.

Not for: Enterprise teams that need process governance, custom field configurations, approval workflows, or team collaboration features like comments and notifications. Tekk is a planning tool, not a project management suite.


What Is Kanban Board Software?

Kanban is a visual workflow method invented at Toyota in the 1940s. Taiichi Ohno adapted lean manufacturing principles into a card-based system for tracking work through stages — the same logic behind every kanban board tool in use today.

In software, a kanban board is a set of columns (typically To Do, In Progress, Done) and cards (individual tasks). Teams move cards left to right as work progresses. The key practice: WIP (work-in-progress) limits. Cap how many cards can be in any column at once, and you force work to finish before new work starts. That simple constraint reduces context-switching and surfaces bottlenecks.

Kanban software ranges from dead-simple (Trello) to enterprise-grade (Jira, Businessmap). Most tools now include AI features: auto-prioritization, bottleneck detection, workload balancing, board generation from documents. What almost none of them do: connect the card to your actual codebase. Tekk.coach was built for that specific gap — codebase-aware AI planning, visualized in a kanban board that stays out of your way.