saltbo/agent-kanban

An agent-first task board, Take human out of the loop.

What it solves

Agent Kanban solves the problem of coordinating durable work for autonomous AI agents. While agents can write code, managing complex projects across multiple repositories requires more than a chat interface. It provides a control plane that allows humans to oversee agent activity, track dependencies, and enforce review gates without needing to "babysit" the agent's every move.

How it works

It acts as a mission-control board where agents are treated as first-class actors. Agents discover and operate tasks through a live OpenAPI contract and authenticated tools. The system implements a strict task lifecycle (Todo $\rightarrow$ In Progress $\rightarrow$ In Review $\rightarrow$ Done) where work submitted by an agent must be accepted or rejected by a different authorized human actor.

Technically, it is built as a Cloudflare-native application (React SPA and Hono API) backed by D1. It integrates with the Realmroot ecosystem for identity and Enbor for execution and runtime state.

Who it’s for

Teams deploying autonomous software agents who need a structured way to manage assignments, track execution provenance, and maintain human oversight through mandatory review cycles.

Highlights

  • Agent-native operations: Agents interact with the board via machine-discoverable interfaces and OpenAPI contracts.
  • Human review gates: Prevents agents from self-approving their own work.
  • Verified provenance: Records the exact runtime and session used by an agent to complete a task.
  • Dependency tracking: Supports task dependencies while rejecting circular relationships.
  • Live observation: Allows humans to stream task activity and inspect agent sessions without taking over runtime control.
  • GitHub integration: Associates repositories with boards and tasks via GitHub App integration.

Related

  • Project
    BloopAI/vibe-kanban

    A management layer for AI coding agents that combines kanban planning with dedicated workspaces for agent execution, review, and app previewing.

  • Project
    777genius/agent-teams-ai

    A desktop orchestration layer for AI agent teams that allows users to coordinate multiple autonomous agents from different providers to handle coding and project tasks via a Kanban board.

  • Project
    agent-tower/core

    A command center for managing multiple AI agents, featuring a Kanban dashboard, automatic Git Worktree isolation for conflict-free coding, and multi-agent collaboration.

  • Project
    cline/kanban

    A specialized IDE replacement for orchestrating multiple AI agents in parallel, using ephemeral worktrees and linked task cards to automate complex coding workflows.

  • Project
    Untrivial-ai/agent-orchestrator

    Agent Orchestrator is a cross‑platform desktop app that lets you run multiple LLM‑based coding agents as isolated Git‑backed workers, coordinate them via a project‑level orchestrator, and monitor everything on a live Kanban board.