tjboudreaux/cc-thinking-skills — reverse-engineered prompt

Build me a Claude Code skill pack that gives the agent a library of structured thinking frameworks instead of just winging it. I want a clean, usable collection of around 39 skills for things like first principles, Bayesian reasoning, systems thinking, OODA, pre mortem, theory of constraints, and other decision making and debugging frameworks. There should be one obvious starting point called thinking model router that helps pick the right skill for the situation, and each skill should be easy to invoke by name inside Claude Code.

Please also include the plugin setup, readable docs, example prompts, and any validation or scoring tools needed to check skill quality. A big part of this should be honesty, not hype, so show the evaluation pipeline clearly and publish the real results even if the skills do not prove a consistent accuracy lift yet. I want it to feel polished, practical, and transparent, something people can install and use right away. Look up current Claude Code plugin docs online if you need to.