Preview — full styling will appear after the next deploy completes.

bmad-method

What is BMAD?

Artifact-driven AI development

BMAD (Breakthrough Method of Agile AI-Driven Development) treats documentation as the primary source of truth. Every AI session starts from a shared artifact chain, not a blank prompt — solving the context amnesia problem that makes unstructured AI development unreliable.

flowchart TD
    P([Problem]) --> BA[BA Persona<br>Analysis]
    BA --> PRD[One-Page PRD]
    PRD --> PM[PM Persona<br>Planning]
    PM --> S[User Stories<br>+ AC]
    S --> AR[Architect Persona<br>Solutioning]
    AR --> AD[Architecture Doc]
    AD --> DV[Developer Persona<br>Implementation]
    DV --> C([Code + Tests])

Most AI coding failures are project management failures. Without persistent documentation, every new session with an AI assistant starts from zero. Decisions made yesterday are forgotten. Architecture choices get relitigated. Features the team explicitly ruled out get re-proposed. The AI isn't being unreliable — it has no context to be reliable with.

BMAD solves this by making documentation — not source code — the primary source of truth. A Product Requirements Document, a set of user stories, and an architecture brief travel with the project. When a developer opens a new AI session, they load the relevant artifacts first. The AI isn't guessing; it's following a specification that a human already reviewed and approved.

The method is built around four phases (Analysis, Planning, Solutioning, Implementation), a set of specialised AI personas (Business Analyst, Product Manager, Architect, Developer, Scrum Master), and a scale-adaptive routing system that matches process depth to project complexity. Small bug fixes go through Quick Flow. Multi-sprint features go through the full artifact chain. The same principles apply at every scale.