sacriphanius/Zhilly — reverse-engineered prompt

Reverse engineered prompt

GitHub

Build me the Zhilly firmware for LilyGO T Embed CC1101 and LilyGO T Watch S3. I want it to feel like a small voice controlled AI cyber assistant that wakes up when I say Nihao Miaoban, talks back naturally, and shows animated eyes or status emotions on the screen.

It should support voice commands for safe authorized lab testing, including RF capture and replay from the SD card, IR record and replay, TV B Gone style IR control, BadUSB typing and DuckyScript from the SD card, LED color and brightness effects, battery checks, sleep, and screen brightness. For the watch version, include network device scanning, simple port scanning, DNS lookup, IR controls, OTA update by URL, and reboot by voice.

Add news requests by voice, keyboard layout switching for the listed languages, and clear confirmations before any disruptive action like jamming or BadUSB. Make it buildable for ESP32 S3 hardware, keep the setup simple, and look up current docs online if you need to.

Want more depth? Deep Reverse