Practitioner resources built from real product work. Free to use, designed to be useful — not comprehensive.
Anti-Sycophancy
Eval Kit
A practitioner's library for catching agreement bias in AI products. Ten evaluation techniques, a scoring rubric, and a primer grounded in recent research. Built for PMs and engineers who ship AI features and need to know if their model is telling users what they want to hear.
Data Pipeline
PO Agent Pack
A self-contained brief for turning any capable language model into a data pipeline product ownership assistant. Thirteen capability cards, a maturity rubric, and a domain primer grounded in SAFe 6 PO responsibilities. Domain-agnostic — works for automotive AI, banking, healthcare, logistics, and others.
More
Resources
More practitioner tools in progress. Check back later.