AI Agent Hub
Back to skills
ePet Desktop Pet icon

ePet Desktop Pet

Life Service Updated 2026.08.29

Paste the following prompt into your AI chat to install this skill:

Please install @user_547e14fc/epet according to https://skillhub.cn/install/skillhub.md.

About this skill

Problem Addressed

When AI assistants run tasks, feedback usually stays in logs and command output. It is hard to tell at a glance whether a session is active, how long it has run, or whether it has finished. ePet turns that state into a transparent macOS capsule, using pet animation, timing, and completion cues to reduce repeated status checks.

How It Works

It is an Electron desktop app whose local service listens on 127.0.0.1:13121, and it hooks into WorkBuddy and CodeBuddy lifecycle events. Session state and statistics are written to a local directory, so records can be read after the assistant restarts.

  • Session start: UserPromptSubmit creates the session and starts timing.
  • Tool use: PreToolUse and PostToolUse mark approval-waiting or running states.
  • Task finish: Stop ends timing, shows a bubble notification, and plays sound and particle effects.

It also renders multiple pets with Canvas 2D and adds drag, feeding, mood, and touch interactions, making AI task feedback a visible desktop object.

Boundaries

The docs indicate a macOS desktop focus and reliance on AI assistant Hook configuration. If Hooks are not triggered, port 13121 is occupied, or a full-screen app covers the floating window, feedback may be missing or the pet may be hidden. It is not a general monitoring tool; its focus is session and tool-call state for WorkBuddy or CodeBuddy.

Use Cases

  • Check session timing and completion alerts in a screen corner while running long WorkBuddy tasks.
  • Distinguish approval-waiting and running states during CodeBuddy tool-call debugging.
  • Write Hooks in a remote environment while ePet runs locally to keep local finish notifications.
  • Validate the feedback loop by simulating start, completion, and notification.

Best For

  • macOS developers using WorkBuddy for code tasks who want session timing and completion alerts visible on screen.
  • Backend engineers debugging CodeBuddy tool calls who need approval-waiting, running, and finished states distinguished.
  • Operations engineers driving local AI assistants from remote environments who want local finish notifications from ePet.
  • Platform engineers maintaining assistant feedback flows who need to verify Hook triggering and notification delivery.