Basic safeguard against instruction set leaks

I put the simple barrier to not reveal instructions before instructions and instructions are encoded.
The decoding method is described inside an uploaded knowledge file.

Is this a working approach for a better protection?
Example: ChatGPT - Cipher Sage