Task Monki

Task Monki is a local desktop app that helps you manage coding tasks across multiple AI agents from a single board. You create a task, choose an agent provider and model, and the tool tracks everything from the initial request through code changes, previews, reviews, and finally a pull request. It's essentially a project manager for AI-assisted development — keeping each task's agent run, diffs, logs, and GitHub status together in one place.

The app runs on your machine and works with several coding agents: Codex, OpenCode, and through ACP compatibility, Cursor Agent, Grok Build, and Claude Agent. You can assign tasks, watch the agent's plan and tool calls in real time, inspect file changes, and launch local previews for each task without repeating setup. A feature called Discourse lets you run multiple agents on the same question — either comparing independent answers or using a lead agent with skeptics and verifiers. The board shows what's running, blocked, or ready for your review.

Task Monki is best for developers who regularly use AI coding agents and want more structure around the workflow. It's especially useful if you juggle several agent-driven tasks at once, need to review and iterate on generated code before shipping, or want to compare how different agents handle the same problem. The alpha is free and MIT-licensed, but the developers warn it's not yet ready for critical repositories — so treat it as a powerful tool for experimentation and non-production work.