BDD, ADR, PRD, WTF: Capturing Decisions for Humans and AI Alike — Michal Cichra, Safe Intelligence
Decision records like ADRs help both humans and LLMs overcome limited context and lost institutional knowledge.
“People forget. LLMs context compact. Humans leave. LLMs have no memory.”
A Safe Intelligence engineer argues that architecture/product decision records (ADR, PRD, BDD) solve a shared weakness of humans and LLMs: limited context and lost rationale over time. He frames machine-enforceable decision docs as a way to give AI agents the 'why' behind code, and briefly notes the launch of Spec 27, a product to test agents. It's a useful engineering practice insight but not a major industry-moving signal.