Why We Systematically Test Every Model and Prompt We Use
Most organizations test LLM prompts with spot-checks and intuition. Three months later, they discover failures in edge cases, formatting errors, and instruction-following problems. Here's why systematic testing matters and how to build evaluation systems that really work.
Strategic Plans Fail Because They're PDFs, Not Practices
I've seen it too many times: strategic plans that sit unopened while teams make decisions by loudest voice. Strategy works when you build three core practices—clear priorities, connections, and progress—into how you work.
Building custom AI workflows for strategy work: What I've learned from Skills
I've been experimenting with Claude Skills for strategy work, and understanding the two types—simple markdown, complex integrations, and portability—makes the difference between useful tools and wasted effort.