AgentBoard is an innovative spatial canvas designed to manage multiple AI coding agents all in one interface. Created by Michael Kulyasov, AgentBoard follows a simple concept: one card equals one conversation. This allows developers to launch agents, drag cards around spatially, and monitor progress at a glance.

Whether you are using Claude Code, Codex, Cursor, or opencode, AgentBoard gives you clear visibility over your coding workflow. Easily see who is currently working, who has finished their task, and who is blocked waiting for your reply. Best of all, AgentBoard is local, free, and open source.

Pros of AgentBoard

Spatial canvas interface where one card equals one conversation
Supports multiple popular AI coding tools including Claude Code, Codex, Cursor, and opencode
Provides clear status tracking to see who is working, finished, or blocked
100% local, free, and open-source software

Cons of AgentBoard

Limited feature set details outside of spatial board management
Requires familiarity with local development tools and AI coding agents
Niche developer-focused utility

Frequently Asked Questions

What is AgentBoard?

AgentBoard is a spatial board and canvas designed for AI coding agents where one card represents one conversation. It allows you to launch agents, organize them spatially, and track their progress.