pingta-guangpingwang

An Electron-based multi-project AI cockpit that orchestrates multiple Claude Code instances with a built-in AI supervisor for one-click dispatch, auto-monitoring, and report aggregation.

14
0
85% credibility
Found May 17, 2026 at 14 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
TypeScript
AI Summary

Claude Harness Desktop is a desktop application that acts like a command center for managing multiple AI coding assistants (Claude Code instances). Instead of juggling several projects yourself, you tell a central 'manager AI' what you want done across all your projects, and it coordinates the work β€” starting terminals, distributing tasks, checking progress, and collecting results β€” all from one friendly window on your screen.

How It Works

1
πŸ“¦ You install the app on your computer

After downloading, you open the app and see a welcoming screen asking you to choose your projects.

2
πŸ“ You pick which projects to manage

You browse your folders and select the software projects you want the AI to help with β€” maybe three at once.

3
πŸ€– The AI manager comes to life

For each project, a separate AI assistant starts working inside its own terminal window, like hiring a team of helpers.

4
πŸ’¬ You tell your manager what you want built

In a simple chat box, you describe what all your projects should do β€” like 'add user login to all three' β€” and the manager coordinates the work.

5
While the AI works, you can:
πŸ‘€
Watch the live feed

See what each project AI is typing and thinking in real-time through colorful terminal output

πŸ“Š
Check progress anytime

A quick status panel shows which projects are busy, which are done, and which need attention

6
πŸ“‹ Each project delivers its work

When an AI finishes a task, you get a neat summary of what changed, with automatic checks to make sure everything looks good.

πŸŽ‰ All your projects are updated together

What would have taken you switching between three windows is now done β€” your manager汇ζŠ₯ everything is complete and ready to use.

Sign up to see the full architecture

5 more

Sign Up Free

Star Growth

See how this repo grew from 14 to 14 stars Sign Up Free
Repurpose This Repo

Repurpose is a Pro feature

Generate ready-to-use prompts for X threads, LinkedIn posts, blog posts, YouTube scripts, and more -- with full repo context baked in.

Unlock Repurpose
AI-Generated Review

What is claude-harness-desktop?

A multi-project AI cockpit built with Electron and TypeScript that lets you control multiple Claude Code instances from a single desktop interface. Instead of switching between terminal windows, you get a centralized dashboard where an AI supervisor agent can dispatch tasks across projects, monitor progress automatically, and aggregate reports. The system uses pseudo-terminals (PTY) to run Claude Code instances in parallel, with built-in tools for task distribution, status checking, and output aggregation. It also includes a plugin ecosystem, workflow automation, and system-level features like tray icons and global hotkeys.

Why is it gaining traction?

The killer feature is the built-in AI supervisor that acts as a manager over your Claude Code workers. You send one instruction, and it coordinates multiple projects, polls for completion, and collects results without manual intervention. The "one-click dispatch" and "report aggregation" capabilities address a real pain point for developers managing related microservices or multi-package repositories. The plugin system and rule engine suggest extensibility beyond the initial use case, and the batch command runner with serial/parallel modes adds practical flexibility for CI/CD-style workflows.

Who should use this?

Backend developers managing multiple related services who want a unified view of AI-assisted development across their stack. DevOps engineers running parallel Claude Code instances for infrastructure-as-code tasks across environments. Teams with monorepos or microservices who need coordinated AI task dispatching without manually orchestrating each terminal. Not recommended for solo developers working on single projectsβ€”the overhead won't pay off.

Verdict

The concept is solid and the architecture shows thoughtful engineering, but the 14-star count and early-stage maturity (v2.5 roadmap, sparse documentation) mean you're an early adopter assuming some risk. The credibility score of 0.8500000238418579% reflects this: functional but unproven at scale. Worth watching, but wait for a 1.0 release or community validation before betting a production workflow on it.

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.