Slash-Command Reference
Every step of the BMad Method is a slash command you launch in the embedded terminal, or, during implementation, by clicking a story phase on Dybs.it. This page is the quick reference: what each command does, who runs it, and where its output goes.
How to run one: open a terminal tab in Dybs.it (or press
Cmd+Kfor the Command Palette), type the command, and pick your LLM provider. The agent takes it from there. During the story cycle, clicking a phase runs the right command for you.
By phase
Phase 1 · Analysis 📊
Understand the problem. → Analysis page
| Command | Does | Output |
|---|---|---|
/bmad-brainstorming | Facilitated ideation through one or many techniques | planning-artifacts |
/bmad-deep-recon | Research to support a decision: market, domain, technical, competitive and more. Run it here, or draft a prompt for your own research tool | planning-artifacts/research |
/bmad-product-brief | Crystallize a concept into a product brief (gentler) | planning-artifacts |
/bmad-prfaq | Working-Backwards PRFAQ that stress-tests the concept | planning-artifacts |
Phase 2 · Planning 📋 🎨
Decide what to build. → Planning page
| Command | Does | Output |
|---|---|---|
/bmad-prd | Create, update, or validate the PRD | planning-artifacts |
/bmad-ux | Produce UX specifications and user journeys | planning-artifacts |
Phase 3 · Solutioning 🏗️
Decide how to build it. → Solutioning page
| Command | Does | Output |
|---|---|---|
/bmad-architecture | Produce the architecture spine (invariants) | planning-artifacts |
/bmad-create-epics-and-stories | Break requirements into epics and stories | planning-artifacts |
Phase 4 · Implementation 💻
Build it, story by story. → Implementation page
| Command | Does | Output |
|---|---|---|
/bmad-sprint-planning | Gate planning readiness, then produce the sprint plan (sprint-status.yaml) | implementation-artifacts |
/bmad-sprint-status | Summarize sprint status and route to next step | (none) |
/bmad-create-story | Create (or validate) the next story's spec | implementation-artifacts |
/bmad-dev-story | Implement the story's tasks and tests | code |
/bmad-code-review | Adversarial code review; loops back to dev if needed | (none) |
/bmad-qa-generate-e2e-tests | Generate E2E / API tests for implemented code | implementation-artifacts |
/bmad-retrospective | Epic-end review: lessons learned, next steps | implementation-artifacts |
/bmad-correct-course | Navigate a significant mid-sprint change | planning-artifacts |
Quality & testing 🧪
Risk-based quality. → Quality & testing page
| Command | Does | Output |
|---|---|---|
/bmad-testarch-test-design | System- or epic-level test plan / strategy | test-artifacts |
/bmad-testarch-atdd | Red-phase acceptance-test scaffolds (ATDD) | test-artifacts |
/bmad-testarch-automate | Expand automated test coverage | test-artifacts |
/bmad-testarch-nfr | Audit non-functional requirements (perf/security/…) | test-artifacts |
/bmad-testarch-trace | Traceability matrix + epic quality-gate decision | test-artifacts |
/bmad-testarch-test-review | Review test quality against best practices | test-artifacts |
/bmad-testarch-ci | Scaffold a CI quality pipeline | test-artifacts |
/bmad-testarch-framework | Initialize Playwright or Cypress | project |
Anytime
Not tied to a single phase.
| Command | Does | Output |
|---|---|---|
/bmad-help | Analyze current state; recommend the next command | (none) |
/bmad-quick-dev | Unified intent-in, code-out: clarify → plan → build → review | implementation-artifacts |
/bmad-project-context | Set up or refresh the agent instructions (AGENTS.md) for an existing codebase | repository root |
/bmad-correct-course | Propose a change when the plan needs to shift | planning-artifacts |
The story-cycle commands, mapped to board phases
When you click a story's phase on Dybs.it, Dybs.it launches the matching command for you:
| Board phase | Icon | Command launched |
|---|---|---|
| Backlog | ○ | /bmad-create-story |
| Ready | ◐ | /bmad-dev-story |
| In Progress | ◑ | /bmad-dev-story |
| Review | ◕ | /bmad-code-review |
| Done | ● | (none, terminal state) |
Development and review loop: if review finds issues, the story returns to
/bmad-dev-story until it's clean. See
the story cycle.
Notes
- Command output locations map to folders under
_bmad-output/(planning-artifacts/,implementation-artifacts/,test-artifacts/) anddocs/(project knowledge). Browse them in the Documents view. - Commands are provider-agnostic. Dybs.it translates each
/bmad-*command for whichever LLM you've chosen, see Terminal & providers. - Not sure what to run next?
/bmad-helpreads your project's current state and recommends the next command.