BMAD Method + Langfuse + Claude Code Agent Teams in Production
· 16 min read
Running AI agents in a real codebase means solving three intertwined problems at once: planning and quality gates (so agents don't drift), observability (so you know what's working), and orchestration (so multiple agents divide work without clobbering each other). In nomadically.work — a remote EU job board with an AI classification and skill-extraction pipeline — these problems are solved by three complementary systems: BMAD v6, Langfuse, and Claude Code Agent Teams. This article explains how each works and how they compose.