*Open Source! , you have the src as well.

  • flow is still 92-93% auto with small issues on doctor phase to rap the task runs other then that

everything works it will plan and run the tasks with multi agents flow.

Working with AI across several software projects can quickly become chaotic. Plans live in one place, project rules in another, health reports somewhere else, and it is often difficult to understand what an agent is doing.

We wanted one clear command center where developers could manage their projects and AI-assisted work without jumping between tools.

Super-Massive, or Suma, gives developers one visual dashboard for all their software projects.

Users can import a project and immediately get a clear overview of what it contains, how healthy it is, and what needs attention.

From each project’s control center, users can:

  • Review project health and important warnings
  • Create and manage development plans
  • Run project checkups with Doctor
  • Organize documentation and project context
  • Configure how AI agents should work
  • Manage agents, workflows, and commands
  • Follow active tasks and progress
  • Review failures and understand what happened

Suma turns complex AI-assisted development into a visible and manageable workflow.

We designed Suma around a simple idea: every important action should be understandable from the interface.

Projects are represented as cards, while each project receives its own cockpit. Longer actions show clear progress, and attention indicators guide users toward the next thing that needs work.

The experience is designed to feel like ground control for software development rather than a collection of disconnected developer tools.

The biggest challenge was presenting a large amount of project information without overwhelming the user.

We also needed to make AI activity transparent. Users should be able to see what is running, what finished, what failed, and what requires their attention.

Another challenge was supporting very different kinds of projects through one consistent interface.

  • Created one dashboard for managing multiple software projects
  • Made project health and problems easy to understand
  • Built a visual control center for plans and AI-agent work
  • Added clear progress for longer-running actions
  • Created guided workflows instead of exposing technical complexity
  • Made Suma capable of managing different types of applications
  • Kept project information organized and portable

AI development tools become much more useful when users can understand and control what they are doing.

The best experience is not just faster automation—it is automation with visible progress, clear boundaries, understandable results, and an obvious next step.

What's next:

Next, we want to make project onboarding even easier, add more ready-made workflows, improve reports, and provide more ways to customize agent teams.

Our goal is to make Suma the everyday command center for building and maintaining software with AI.

Built With

  • docker
  • docker-compose
  • fastapi
  • graphify
  • lore
  • model-context-protocol
  • motion
  • openai-codex
  • pydantic
  • pytest
  • python
  • radix-ui
  • react
  • server-sent-events
  • shadcn/ui
  • sqlite
  • tailwind-css
  • typescript
  • uvicorn
  • vite
  • vitest
Share this project:

Updates