CheswickDEV/claude-opus-4.8-prompt-optimizer ? reverse-engineered prompt
Reverse engineered prompt
Build me a Claude Opus 4.8 prompt optimizer that takes a messy raw prompt and turns it into a polished, production ready prompt with XML sections.
I want it to follow Anthropic style best practices, keep the user’s original language, and adjust how much it rewrites based on how simple or complex the input is. For simple prompts, keep it light, for detailed prompts, add more structure, clearer constraints, and a better output format. Make sure it only runs when the input includes the hard prompt: trigger, and that it gives a clean optimized prompt I can paste straight into Claude.
Please include the main workflow, prompt analysis, routing, structure, optimization, and a final quality check. Also add a clear quick start guide and examples showing a before and after transformation. If you need to check current Claude docs online to match the latest guidance, do that.
Are you gonna build this?
make sure you review the code using coderabbit