kandev
AI Kanban & Development Environment. Orchestrate multiple agents, review changes, open PRs. Multi-provider, self-hostable, no telemetry.
What is it?
What it is
An open-source agentic development environment and orchestrator that allows users to manage, run, and review AI agent tasks in parallel using kanban and pipeline views.
Why it exists
To provide a human-in-the-loop control plane for AI agents that solves the limitations of terminal-based agent interaction, offering a centralized workspace for reviewing changes, managing multi-provider agents, and offloading execution to remote runtimes.
Who should use it
Developers working with AI agents, teams needing parallel task execution and workflow automation, self-hosted solutions users
Who should avoid it
Users who don't require agent orchestration, prefer cloud-only solutions, or need simple task management
How it works
A quick walkthrough in plain English
How kandev works
Step 1 of 3
You interact with it
Open kandev, send a request, or connect it to your stack.
Features
Advantages
- Open source and self‑hostable, no telemetry
- Multi‑provider agent ecosystem, no vendor lock‑in
- Server‑first architecture allows access from any device
- Remote runtimes enable scaling beyond local machine
- Integrated workspace reduces context switching
- Parallel execution boosts productivity
- Automated workflow and Kanban management
- Review‑first design ensures human oversight
- Customizable runtime profiles and prompts
- Extensible via external MCP and agent adapters
- Portable workflows across installations
- Statistical insights for productivity
- CLI passthrough preserves native TUI experience
Disadvantages
- Requires setup of executors (Docker, SSH, etc.)
- Learning curve for configuring agents and workflows
- AGPL license may restrict commercial use
- Limited built‑in CI/CD integration
- Performance overhead of server and worktree management
- Potential network latency for remote runtimes
- Office mode still in progress
- Requires Node.js and Go toolchain for development
Installation
cloud
Runs the agent in a remote cloud environment via sprites.dev
docker
Runs the agent in an isolated Docker container
native
brew install kdlbs/kandev/kandev kandev OR npx kandev@latest OR npm install -g kandev@latest kandev
FAQ
Which AI agents can I use with Kandev?
Kandev supports a wide range of agents via the Agent Client Protocol (ACP), including Claude Code, GitHub Copilot, Gemini CLI, Cursor, Devin, Qwen, and many others. You can also run any agent CLI in TUI mode using CLI passthrough.
How does Kandev prevent agents from conflicting with each other?
Kandev uses Git worktrees to provide workspace isolation, ensuring that concurrent agents working on different tasks do not interfere with each other's file changes.
Can I run agents on a remote machine instead of my local computer?
Yes. Kandev supports multiple executors, allowing you to run agents as local processes, in isolated Docker containers, on remote servers via SSH, or in cloud environments like sprites.dev.
How do I manage complex, multi-step AI workflows?
You can define agentic workflows that mix and match different agents for different steps—for example, using one agent for planning and another for implementation—and manage them through Kanban or pipeline views.
Is my data private and can I host it myself?
Yes. Kandev is open-source, contains no telemetry, and is designed to be self-hosted on your own infrastructure, allowing you to access it securely via Tailscale or a VPN.
Featured in Videos
YouTube tutorials and walkthroughs for kandev
Alternatives
Similar projects ranked by category, topics, and text overlap.
