WebMCP Agent Kanban

Manage project work through UI or AI tools.

Checking WebMCP…
Workspace / Agent platform

From prototype to production

Platform engineering · Sprint 01 / A shared workspace for humans & agents

One board. Two ways to build.
Drag to move · Click or press Enter to edit · Changes saved locally

Activity Log

A transparent trail across both capability paths.

● LIVE

Project Signals

Derived from the shared board state. Always current.

Try asking your AI agent

Click to copy a prompt into your connected agent.

⌘   WebMCP Tool Inspector

Structured capabilities, not DOM automation.

25 tools · Expand ↓
Checking browser capabilities…
Capabilities, not clicks

How WebMCP powers this board

  1. The page exposes structured project-management capabilities.
  2. The AI agent discovers those capabilities.
  3. The agent reads project state when needed.
  4. The agent chooses appropriate tools.
  5. BoardEngine updates the shared application state.
  6. The same Kanban UI immediately reflects those changes.
Traditional browser agent
AI → inspect DOM → find card → find button → click → wait → inspect again
WebMCP agent
AI → board_search_tasks() → board_batch_move()
AI Agent
↓
WebMCP Tools
↓
Human User
↓
UI Controls / Drag & Drop
↓
BoardEngine
↓
Shared Board State
↓
Kanban UI

AI and humans use the same underlying capability layer.

No backend. No simulated clicks. Real application state.