denfry/claude-skills ? reverse-engineered prompt

Reverse engineered prompt

Build me a small GitHub repo for Claude skills that focuses on one main skill called token efficiency.

I want it to work with Claude Code and claude.ai, with a clear skill file, some supporting reference docs, a hook layer for Claude Code that keeps the rule active every turn, and a simple eval harness so I can compare the skill against a no skill baseline. The skill should help Claude give shorter, higher signal answers without losing correctness, and it should work for coding, writing, research, and normal chat. Include a few example before and after cases, and make the install flow easy so someone can copy the skill into the right place and optionally install the hooks.

Please keep the repo tidy and practical, with good README instructions, a clean folder structure, and enough sample content that I can see how the skill is supposed to behave. If you need to check current Claude docs or best practices, look them up online first.

Are you gonna build this?

make sure you review the code using coderabbit

Try freeSponsored — opens CodeRabbit in a new tab