nailong-pet-skill
Adds Nailong, a popular Chinese internet meme character, as an animated desktop pet to OpenAI Codex, and doubles as a template for adding any custom character with its own sprite sheet.
This repository adds a custom animated desktop pet to OpenAI Codex, which is an AI-powered coding assistant with a visual interface. Codex has a built-in feature where a small animated character can sit on your screen while you work, and this project lets you swap in a character called Nailong, a popular internet meme character from Chinese social media.
The repository contains the animation sprite sheet (a single image file where all the character's frames are laid out in a grid), a preview image, and a Python installation script. Running the script copies the image and a configuration file into Codex's pet folder. After restarting Codex, the Nailong character appears as an option in the settings under the pets section.
The character has nine animation states that correspond to what Codex is doing at any moment: standing still when idle, running left or right when you drag it, waving when it finishes a task, crouching when something fails, and tilting its head with a question mark during code review, among others.
The project is written in Chinese and also functions as a template for anyone who wants to add a different custom character to Codex. The README walks through how to prepare your own sprite sheet (which must follow a specific grid layout of 8 columns by 9 rows at fixed pixel dimensions), swap in your own images, and update the configuration files to use your character's name and trigger words. Installation can also be done through a command-line tool called skills.sh with a single command.
Where it fits
- Replace the default Codex desktop pet with the Nailong meme character using the one-command install script.
- Use this repository as a template to create your own custom animated desktop pet for Codex with any character you design.
- Learn how Codex's pet system works, sprite sheet format, configuration files, and animation state triggers, so you can build on it.