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

bmad-method

The Four Phases

Analysis → Planning → Solutioning → Implementation

BMAD organises every project into four sequential phases, each producing specific artifacts that feed the next. Work cannot begin in a phase until the prior phase's artifacts exist and have been reviewed by a human.

flowchart LR
    A([Analysis]) --> AP[One-Page PRD]
    AP --> PL([Planning])
    PL --> PS[User Stories<br>+ AC]
    PS --> SO([Solutioning])
    SO --> AD[Architecture<br>Doc + ADRs]
    AD --> IM([Implementation])
    IM --> CO[Code + Tests]

The four phases are Analysis, Planning, Solutioning, and Implementation. They are sequential by design: each phase produces artifacts that the next phase depends on. Skipping a phase doesn't save time — it transfers unresolved decisions into a later phase where they are more expensive to fix.

Analysis produces a one-page PRD capturing the problem, constraints, success criteria, and the users affected. Planning takes the PRD and produces user stories with acceptance criteria, prioritised against the actual constraint set. Solutioning produces the minimum architectural decisions needed to start implementation without painting yourself into a corner. Implementation works through the stories iteratively, with the AI producing code against the specifications already decided.

The artifact chain is the mechanism that makes BMAD work. When each phase's output is reviewed and approved before the next phase begins, decisions are made at the right level of abstraction. An architecture constraint caught at the Solutioning phase costs far less than one discovered mid-implementation. Each handoff is a gate, not a formality.