witnesstodark/mr-mak-workspace

A local desktop workspace for Codex and Claude Code, with project reports, creative skills and optional voice.

What it solves

Mr. Mak Workspace provides a unified visual interface for managing AI agents that typically operate in a command-line interface (CLI). It bridges the gap between raw terminal output and a structured workspace where users can organize research, images, reports, and project files alongside their AI conversations.

How it works

The system consists of two primary desktop windows:

  • Chats: A managed terminal environment where users interact with agents like Codex CLI, Claude Code, or OpenCode. Each conversation is saved with its own history and tab.
  • Workspace: A visual area for reviewing results, managing project cards, browsing files, and organizing knowledge and processes.

It integrates with existing CLI agent installations and uses the user's own accounts and subscriptions. It also features a mobile web app (connected via Tailscale) that allows users to follow chats, send messages, and use voice dictation while the agents continue to run on the computer.

Who it’s for

Developers, researchers, and creative professionals (such as those working in Blender or game engines) who use CLI-based AI agents and want a visual way to organize their project assets and agent outputs.

Highlights

  • Multi-Agent Support: Integrates with Codex CLI, Claude Code, and OpenCode.
  • Cross-Device Sync: Mobile access via Tailscale for remote monitoring and voice input.
  • Visual Project Management: Uses cards and tabs to organize research, design versions, and motion tests.
  • Integrated File Browser: Direct access to project folders, Markdown notes, and media previews.
  • Extensible Skills: Includes a library of project skills for tasks like VFX, UI, and 3D modeling.
  • MCP Support: Visibility and management of Model Context Protocol (MCP) connections.

Related

  • Project
    kydlikebtc/awesome-jev

    awesome-jev is a curated catalog of public code examples, tutorials, and benchmarks that show how to use TypeSafe AI’s Jev decision‑model service (choice, score, noul primitives). It organizes over 1,200 resources by decision pattern, provides a primitive‑picker guide, and includes independent measurement reports (both positive and negative). The repo itself contains no SDK or model code; it only indexes and documents existing usage.

  • Project
    agentvivekkumar/dontbemichael

    A desktop application that manages a team of AI agents to handle small business operations like email, scheduling, and administrative tasks, with all work running locally on your machine.

  • Project
    sam70361/aora-bot

    A zero-dependency SVG and JavaScript expression engine that provides 32 emotional states for AI assistants, chatbots, and desktop pets.

  • Project
    selfimproving-agent/Awesome-Self-Improving-Agents

    This is a curated, continuously evolving hub that collects papers, benchmarks, blogs, podcasts, and other resources on self-improving agentic systems. It organizes the field into two main branches—improving the underlying model and improving the surrounding scaffolding—and serves as a practical guide for researchers and practitioners.