BMad Method is a free and open-source method for agile, AI-driven development: install it with npx, invoke bmad-build with the change you want, and it right-sizes the process — straight to implementation for small changes, deeper planning with product, architecture, UX, development and testing perspectives for larger ones. It is a strong answer to 'how does this change get delivered.' It is not an answer to 'what does my AI already know about me, and what did I ask for last week.'
BMad Method: a free, open-source, MIT-licensed method for Agile AI-Driven Development, installed with `npx bmad-method install`. You invoke `bmad-build` with what you want changed; the method sizes itself to the work, brings in specialized product, architecture, UX, development and testing perspectives when they help, and preserves the decisions as context for the work that follows. There are no paywalled workflows.
Brainfile: a maintained operating layer for Claude Code — CLAUDE.md plus brain/, an Autopilot run protocol, a never-empty backlog, a directive ledger that tracks every ask until proven done, memory, and an end-of-run brief. It is read at every session start, whether or not you are delivering software.
Both care about durable context, which is why they get compared. The difference is scope: BMad's context is the DECISION TRAIL OF THE WORK — briefs, specs, architecture, carried forward across the delivery. Brainfile's context is THE OPERATOR — who you are, how you want work run, and every ask you have made, carried across all sessions including the ones with no delivery in them.
| Feature | BMad Method | Brainfile |
|---|---|---|
| Unit of work | A change: idea or request → working software | The operator: standing context, protocol and memory |
| What persists | Briefs, specs and architecture for the work | Role, constraints, prior decisions and every ask made |
| Specialized perspectives | Product, architecture, UX, development, testing | 15 role packs plus your own generated role context |
| Applies outside delivery | Aimed at building software | Applies to research, ops, review and admin sessions too |
| Invocation | You invoke the workflow when you want it | Automatic — loaded at every session start |
| Tracks an ask until done | Explicit decisions for the change in flight | Directive ledger tracks every ask until it is PROVEN done |
| Prerequisites | Node.js, Python and uv | None beyond your existing Claude Code setup |
| Reporting | Artifacts and the decision trail | End-of-run brief: accomplished / needs-you / next priorities |
| Extensibility | Community project with an active ecosystem | 10 pattern files, 15 role packs, weekly freshness feed |
| Price | Free — open source, no paid tier to unlock the workflow | Paid subscription — Brainfile Pro, with a 30-day free trial and no card |
You want a real method for turning requests into shipped software without losing the thinking — right-sized planning, explicit decisions, and specialist perspectives on demand. It is free, MIT-licensed, and there is nothing gated behind a paywall.
The method is not your bottleneck; the amnesia is. Your AI restarts every session with no idea who you are, what you have ruled out, or what you asked for on Tuesday, and it never reports back what it did. Brainfile is the layer that fixes that, and it applies to sessions that are not deliveries.
They stack cleanly. BMad governs the delivery of a change; Brainfile governs the operator that runs it and remembers across changes. A CLAUDE.md carrying your role, constraints and prior decisions makes BMad's clarify and plan phases shorter, because the assumptions it would otherwise have to surface are already written down.
BMad Method is a process you enter. It sizes itself to the work, makes the important decisions explicit rather than letting an assistant assume them, and preserves that decision trail for the next change. Its value is concentrated in the moment you are delivering something.
Brainfile is not something you enter. It is what is already true when the session opens: your role, your rules, your backlog, the ledger of every ask, and the protocol the run will follow. It is worth most in the gaps between deliveries, where a method has nothing to say.
Brainfile Pro is maintained operating context, a run protocol and a directive ledger your AI reads every session. 30-day free trial, no card.
30-day money-back guarantee · Instant access · Cancel anytime
AI context engineering tips and new configuration releases. Free.
No spam. Unsubscribe anytime.