RS

ringhyacinth/star-office-ui

Developer tools

This page was translated automatically. Technical details may differ from the original. View the original

7.5K stars Quality 53 Trend 31NOASSERTION

A pixel office for your OpenClaw: turn invisible work states into a cozy little space with characters, daily notes, and guest agents. Code under MIT; art assets for non-commercial learning only.

Overview

— Visualizes the AI assistant's working status in real time, letting you directly see "who is doing what, what was done yesterday, and whether it is online now." Supports multi-agent collaboration, trilingual Chinese, English, and Japanese, AI-generated image decoration, and desktop pet mode. Best experience when deeply integrated with OpenClaw, can also be deployed independently as a status dashboard. This project is co-created by and continues to maintain and co-build with community developers (@Zhaohan-Wang, @Jah-yee, @liaoandi). Submissions of Issues and PRs are welcome, and we also thank every contributor for their support. Lobster will automatically complete clone, install dependencies, start the backend, synchronize configuration status, and send you the access address. (The code uses X | Y union type syntax and does not support versions 3.9 or earlier) This is **. The Agent automatically switches states while working, and pixel characters in the office will walk to the corresponding area in real time—you just need to open the webpage to see what the AI is doing at that moment. It can also be fully deployed. You can: - Manually or script push status via set_state.py or API - Treat it as a pixel-style personal status page or remote work dashboard - Connect to any system capable of sending HTTP requests to drive status 1. —— Automatically maps 6 states (idle / writing / research / executing / syncing / error) to different office areas, with animation + bubbles displayed in real time 2. —— Automatically reads the most recent day's work records from memory/*.md, displays them as "Yesterday's Notes" card after desensitization 3. —— Invite other Agents to your office via join key to view multiple user statuses in real time 4. —— One-click switching between CN / EN / JP, interface copy, bubbles, and loading prompts all synced 5. —— Sidebar manages characters/scenes/decorative materials, supports dynamic frame synchronization to avoid flicker 6. —— Integrates Gemini API to use AI to change office backgrounds; Core functions can be used normally without API connection 7. —— Open directly on phone to view, perfect for quick glances when out and out 8. —— Sidebar password protection, weak password interception in production environments, session cookie hardening 9. —— Recommended Cloudflare Tunnel for one-click public networking, also using your own domain name / reverse proxy 10. —— Optional Electron desktop packaging turns the office into a desktop pet for transparent windows (see description below)

Overview (machine translated)

🌐 Language: Chinese | English | Japanese

A pixel-style AI office board — visualizes the AI assistant’s work status in real time, letting you directly see “who is doing what, what was done yesterday, and whether it is online now.”

Supports multi-agent collaboration, trilingual Chinese, English, and Japanese, AI-generated image decoration, and desktop pet mode. The best experience is when deeply integrated with OpenClaw, and it can also be deployed independently as a status board.

This project was co-created by Ring Hyacinth and Simon Lee, and continuously maintained and co-built with community developers (@Zhaohan-Wang, @Jah-yee, @liaoandi). Issue and PR submissions are welcome, and we appreciate every contributor’s support.


✨ Quick Experience

If you are using OpenClaw, send the following message directly to your lobster:

Lobster will automatically complete clone, install dependencies, start the backend, synchronize configuration status, and send you the access address.

Method 2: Manual deployment in 30 seconds

Environment requirements: Python 3.10+ (code uses [T0]] union type syntax, does not support versions 3.9 and below)

Open http://127.0.0.1:19000 and try switching the state:


🤔 Who is it suitable for?

Users with OpenClaw / AI Agent

This is the complete experience. The Agent automatically switches modes while working, and pixel characters in the office will walk to the corresponding area in real time—you just need to open the webpage to see what the AI is doing at that moment.

Users without OpenClaw

It can also be fully deployed. You can:

  • Manually or script push status using set_state.py or API
  • Treat it as a pixel-style personal status page / remote work dashboard
  • Connect any system capable of sending HTTP requests to drive state

📋 Overview of features

  1. State Visualization — Automatically maps 6 states (idle / writing / researching / executing / syncing / error) to different areas of the office, with animation + bubbles displayed in real time
  2. Yesterday’s Notes — Automatically reads the most recent day’s work records from memory/*.md, and after desensitization, displays them as “Yesterday’s Notes” cards
  3. Multi-Agent Collaboration — Invite other agents to join your office via join key and check the status of multiple participants in real time
  4. Chinese, English, Japanese — One-click switching between CN / EN / JP, with interface text, bubbles, and loading prompts all linked
  5. Custom Art Assets — Sidebar management for characters, scenes, and decorative materials, supports dynamic frame synchronization to avoid flickering

This page was translated automatically. Technical details may differ from the original.

Recommended Tools

Try a different keyword or remove a filter.

Install

npx skillfish add ringhyacinth/star-office-ui
New to MCP? Start with the mcpskill.org guide →