BMad Method — User Guide

This guide will help you understand and effectively use the BMad Method for agile AI-driven planning and development.

The BMad Plan and Execute Workflow

First, here is the full standard Greenfield Planning + Execution Workflow. Brownfield is very similar, but it's suggested to understand this greenfield first, even if on a simple project before tackling a brownfield project. The BMad Method needs to be installed to the root of your new project folder. For the planning phase, you can optionally perform it with powerful web agents, potentially resulting in higher quality results at a fraction of the cost it would take to complete if providing your own API key or credits in some Agentic tools. For planning, powerful thinking models and larger context - along with working as a partner with the agents will net the best results.

If you are going to use the BMad Method with a Brownfield project (an existing project), review Working in the Brownfield.

If the diagrams below don't render, install Markdown All in One along with the Markdown Preview Mermaid Support plugins to VSCode (or one of the forked clones). With these plugins, if you right click on the tab when open, there should be an Open Preview option, or check the IDE documentation.

The Planning Workflow (Web UI or Powerful IDE Agents)

Before development begins, BMad follows a structured planning workflow that's ideally done in web UI for cost efficiency:

graph TD
    A["Start: Project Idea"] --> B{"Optional: Analyst Research"}
    B -->|Yes| C["Analyst: Brainstorming (Optional)"]
    B -->|No| G{"Project Brief Available?"}
    C --> C2["Analyst: Market Research (Optional)"]
    C2 --> C3["Analyst: Competitor Analysis (Optional)"]
    C3 --> D["Analyst: Create Project Brief"]
    D --> G
    G -->|Yes| E["PM: Create PRD from Brief (Fast Track)"]
    G -->|No| E2["PM: Interactive PRD Creation (More Questions)"]
    E --> F["PRD Created with FRs, NFRs, Epics & Stories"]
    E2 --> F
    F --> F2{"UX Required?"}
    F2 -->|Yes| F3["UX Expert: Create Front End Spec"]
    F2 -->|No| H["Architect: Create Architecture from PRD"]
    F3 --> F4["UX Expert: Generate UI Prompt for Lovable/V0 (Optional)"]
    F4 --> H2["Architect: Create Architecture from PRD + UX Spec"]
    H --> Q{"Early Test Strategy? (Optional)"}
    H2 --> Q
    Q -->|Yes| R["QA: Early Test Architecture Input on High-Risk Areas"]
    Q -->|No| I
    R --> I["PO: Run Master Checklist"]
    I --> J{"Documents Aligned?"}
    J -->|Yes| K["Planning Complete"]
    J -->|No| L["PO: Update Epics & Stories"]
    L --> M["Update PRD/Architecture as needed"]
    M --> I
    K --> N["📁 Switch to IDE (If in a Web Agent Platform)"]
    N --> O["PO: Shard Documents"]
    O --> P["Ready for SM/Dev Cycle"]

    style A fill:#f5f5f5,color:#000
    style B fill:#e3f2fd,color:#000
    style C fill:#e8f5e9,color:#000
    style C2 fill:#e8f5e9,color:#000
    style C3 fill:#e8f5e9,color:#000
    style D fill:#e8f5e9,color:#000
    style E fill:#fff3e0,color:#000
    style E2 fill:#fff3e0,color:#000
    style F fill:#fff3e0,color:#000
    style F2 fill:#e3f2fd,color:#000
    style F3 fill:#e1f5fe,color:#000
    style F4 fill:#e1f5fe,color:#000
    style G fill:#e3f2fd,color:#000
    style H fill:#f3e5f5,color:#000
    style H2 fill:#f3e5f5,color:#000
    style Q fill:#e3f2fd,color:#000
    style R fill:#ffd54f,color:#000
    style I fill:#f9ab00,color:#fff
    style J fill:#e3f2fd,color:#000
    style K fill:#34a853,color:#fff
    style L fill:#f9ab00,color:#fff
    style M fill:#fff3e0,color:#000
    style N fill:#1a73e8,color:#fff
    style O fill:#f9ab00,color:#fff
    style P fill:#34a853,color:#fff

Web UI to IDE Transition

Critical Transition Point: Once the PO confirms document alignment, you must switch from web UI to IDE to begin the development workflow:

Copy Documents to Project: Ensure docs/prd.md and docs/architecture.md are in your project's docs folder (or a custom location you can specify during installation)
Switch to IDE: Open your project in your preferred Agentic IDE
Document Sharding: Use the PO agent to shard the PRD and then the Architecture
Begin Development: Start the Core Development Cycle that follows

Planning Artifacts (Standard Paths)

PRD              → docs/prd.md
Architecture     → docs/architecture.md
Sharded Epics    → docs/epics/
Sharded Stories  → docs/stories/
QA Assessments   → docs/qa/assessments/
QA Gates         → docs/qa/gates/

The Core Development Cycle (IDE)

Once planning is complete and documents are sharded, BMad follows a structured development workflow:

graph TD
    A["Development Phase Start"] --> B["SM: Reviews Previous Story Dev/QA Notes"]
    B --> B2["SM: Drafts Next Story from Sharded Epic + Architecture"]
    B2 --> S{"High-Risk Story? (Optional)"}
    S -->|Yes| T["QA: *risk + *design on Draft Story"]
    S -->|No| B3
    T --> U["Test Strategy & Risk Profile Created"]
    U --> B3{"PO: Validate Story Draft (Optional)"}
    B3 -->|Validation Requested| B4["PO: Validate Story Against Artifacts"]
    B3 -->|Skip Validation| C{"User Approval"}
    B4 --> C
    C -->|Approved| D["Dev: Sequential Task Execution"]
    C -->|Needs Changes| B2
    D --> E["Dev: Implement Tasks + Tests"]
    E --> V{"Mid-Dev QA Check? (Optional)"}
    V -->|Yes| W["QA: *trace or *nfr for Early Validation"]
    V -->|No| F
    W --> X["Dev: Address Coverage/NFR Gaps"]
    X --> F["Dev: Run All Validations"]
    F --> G["Dev: Mark Ready for Review + Add Notes"]
    G --> H{"User Verification"}
    H -->|Request QA Review| I["QA: Test Architect Review + Quality Gate"]
    H -->|Approve Without QA| M["IMPORTANT: Verify All Regression Tests and Linting are Passing"]
    I --> J["QA: Test Architecture Analysis + Active Refactoring"]
    J --> L{"QA Decision"}
    L -->|Needs Dev Work| D
    L -->|Approved| M
    H -->|Needs Fixes| D
    M --> N["IMPORTANT: COMMIT YOUR CHANGES BEFORE PROCEEDING!"]
    N --> Y{"Gate Update Needed?"}
    Y -->|Yes| Z["QA: *gate to Update Status"]
    Y -->|No| K
    Z --> K["Mark Story as Done"]
    K --> B

    style A fill:#f5f5f5,color:#000
    style B fill:#e8f5e9,color:#000
    style B2 fill:#e8f5e9,color:#000
    style S fill:#e3f2fd,color:#000
    style T fill:#ffd54f,color:#000
    style U fill:#ffd54f,color:#000
    style B3 fill:#e3f2fd,color:#000
    style B4 fill:#fce4ec,color:#000
    style C fill:#e3f2fd,color:#000
    style D fill:#e3f2fd,color:#000
    style E fill:#e3f2fd,color:#000
    style V fill:#e3f2fd,color:#000
    style W fill:#ffd54f,color:#000
    style X fill:#e3f2fd,color:#000
    style F fill:#e3f2fd,color:#000
    style G fill:#e3f2fd,color:#000
    style H fill:#e3f2fd,color:#000
    style I fill:#f9ab00,color:#fff
    style J fill:#ffd54f,color:#000
    style K fill:#34a853,color:#fff
    style L fill:#e3f2fd,color:#000
    style M fill:#ff5722,color:#fff
    style N fill:#d32f2f,color:#fff
    style Y fill:#e3f2fd,color:#000
    style Z fill:#ffd54f,color:#000

Prerequisites

Before installing BMad Method, ensure you have:

Node.js ≥ 18, npm ≥ 9
Git installed and configured
(Optional) VS Code with "Markdown All in One" + "Markdown Preview Mermaid Support" extensions

Installation

Optional

If you want to do the planning on the web with Claude (Sonnet 4 or Opus), Gemini Gem (2.5 Pro), or Custom GPTs:

Navigate to dist/teams/
Copy team-fullstack.txt
Create new Gemini Gem or CustomGPT
Upload file with instructions: "Your critical operating instructions are attached, do not break character as directed"
Type /help to see available commands

IDE Project Setup

# Interactive installation (recommended)
npx bmad-method install

Codex (CLI & Web)

BMAD integrates with OpenAI Codex via AGENTS.md and committed core agent files.

Two installation modes:
- Codex (local only): keeps .bmad-core/ ignored for local dev.
- npx bmad-method install -f -i codex -d .
- Codex Web Enabled: ensures .bmad-core/ is tracked so you can commit it for Codex Web.
- npx bmad-method install -f -i codex-web -d .
What gets generated:
- AGENTS.md at the project root with a BMAD section containing
- How-to-use with Codex (CLI & Web)
- Agent Directory (Title, ID, When To Use)
- Detailed per‑agent sections with source path, when-to-use, activation phrasing, and YAML
- Tasks with quick usage notes
- If a package.json exists, helpful scripts are added:
- bmad:refresh, bmad:list, bmad:validate
Using Codex:
- CLI: run codex in the project root and prompt naturally, e.g., “As dev, implement …”.
- Web: commit .bmad-core/ and AGENTS.md, then open the repo in Codex and prompt the same way.
Refresh after changes:
- Re-run the appropriate install mode (codex or codex-web) to update the BMAD block in AGENTS.md.

Special Agents

There are two BMad agents — in the future they'll be consolidated into a single BMad-Master.

BMad-Master

This agent can do any task or command that all other agents can do, aside from actual story implementation. Additionally, this agent can help explain the BMad Method when on the web by accessing the knowledge base and explaining anything to you about the process.

If you don't want to bother switching between different agents aside from the dev, this is the agent for you. Just remember that as the context grows, the performance of the agent degrades, therefore it is important to instruct the agent to compact the conversation and start a new conversation with the compacted conversation as the initial message. Do this often, preferably after each story is implemented.

BMad-Orchestrator

This agent should NOT be used within the IDE, it is a heavyweight, special-purpose agent that utilizes a lot of context and can morph into any other agent. This exists solely to facilitate the teams within the web bundles. If you use a web bundle you will be greeted by the BMad Orchestrator.

How Agents Work

Dependencies System

Each agent has a YAML section that defines its dependencies:

dependencies:
  templates:
    - prd-template.md
    - user-story-template.md
  tasks:
    - create-doc.md
    - shard-doc.md
  data:
    - bmad-kb.md

Key Points:

Agents only load resources they need (lean context)
Dependencies are automatically resolved during bundling
Resources are shared across agents to maintain consistency

Agent Interaction

In IDE:

# Some IDEs, like Cursor or Windsurf for example, utilize manual rules so interaction is done with the '@' symbol
@pm Create a PRD for a task management app
@architect Design the system architecture
@dev Implement the user authentication

# Some IDEs, like Claude Code, use slash commands instead
/pm Create user stories
/dev Fix the login bug

Interactive Modes

Incremental Mode: Step-by-step with user input
YOLO Mode: Rapid generation with minimal interaction

IDE Integration

IDE Best Practices

Context Management: Keep relevant files only in context, keep files as lean and focused as necessary
Agent Selection: Use appropriate agent for task
Iterative Development: Work in small, focused tasks
File Organization: Maintain clean project structure
Commit Regularly: Save your work frequently

The Test Architect (QA Agent)

Overview

The QA agent in BMad is not just a "senior developer reviewer" - it's a Test Architect with deep expertise in test strategy, quality gates, and risk-based testing. Named Quinn, this agent provides advisory authority on quality matters while actively improving code when safe to do so.

Quick Start (Essential Commands)

@qa *risk {story}       # Assess risks before development
@qa *design {story}     # Create test strategy
@qa *trace {story}      # Verify test coverage during dev
@qa *nfr {story}        # Check quality attributes
@qa *review {story}     # Full assessment → writes gate

Command Aliases (Test Architect)

The documentation uses short forms for convenience. Both styles are valid:

*risk    → *risk-profile
*design  → *test-design
*nfr     → *nfr-assess
*trace   → *trace-requirements (or just *trace)
*review  → *review
*gate    → *gate

Core Capabilities

1. Risk Profiling (`*risk`)

When: After story draft, before development begins (earliest intervention point)

Identifies and assesses implementation risks:

Categories: Technical, Security, Performance, Data, Business, Operational
Scoring: Probability × Impact analysis (1-9 scale)
Mitigation: Specific strategies for each identified risk
Gate Impact: Risks ≥9 trigger FAIL, ≥6 trigger CONCERNS (see tasks/risk-profile.md for authoritative rules)

2. Test Design (`*design`)

When: After story draft, before development begins (guides what tests to write)

Creates comprehensive test strategies including:

Test scenarios for each acceptance criterion
Appropriate test level recommendations (unit vs integration vs E2E)
Risk-based prioritization (P0/P1/P2)
Test data requirements and mock strategies
Execution strategies for CI/CD integration

Example output:

test_summary:
  total: 24
  by_level:
    unit: 15
    integration: 7
    e2e: 2
  by_priority:
    P0: 8 # Must have - linked to critical risks
    P1: 10 # Should have - medium risks
    P2: 6 # Nice to have - low risks

3. Requirements Tracing (`*trace`)

When: During development (mid-implementation checkpoint)

Maps requirements to test coverage:

Documents which tests validate each acceptance criterion
Uses Given-When-Then for clarity (documentation only, not BDD code)
Identifies coverage gaps with severity ratings
Creates traceability matrix for audit purposes

4. NFR Assessment (`*nfr`)

When: During development or early review (validate quality attributes)

Validates non-functional requirements:

Core Four: Security, Performance, Reliability, Maintainability
Evidence-Based: Looks for actual implementation proof
Gate Integration: NFR failures directly impact quality gates

5. Comprehensive Test Architecture Review (`*review`)

When: After development complete, story marked "Ready for Review"

When you run @qa *review {story}, Quinn performs:

Requirements Traceability: Maps every acceptance criterion to its validating tests
Test Level Analysis: Ensures appropriate testing at unit, integration, and E2E levels
Coverage Assessment: Identifies gaps and redundant test coverage
Active Refactoring: Improves code quality directly when safe
Quality Gate Decision: Issues PASS/CONCERNS/FAIL status based on findings

6. Quality Gates (`*gate`)

When: After review fixes or when gate status needs updating

Manages quality gate decisions:

Deterministic Rules: Clear criteria for PASS/CONCERNS/FAIL
Parallel Authority: QA owns gate files in docs/qa/gates/
Advisory Nature: Provides recommendations, not blocks
Waiver Support: Documents accepted risks when needed

Note: Gates are advisory; teams choose their quality bar. WAIVED requires reason, approver, and expiry date. See templates/qa-gate-tmpl.yaml for schema and tasks/review-story.md (gate rules) and tasks/risk-profile.md for scoring.

Working with the Test Architect

Integration with BMad Workflow

The Test Architect provides value throughout the entire development lifecycle. Here's when and how to leverage each capability:

Stage	Command	When to Use	Value	Output
Story Drafting	`*risk`	After SM drafts story	Identify pitfalls early	`docs/qa/assessments/{epic}.{story}-risk-{YYYYMMDD}.md`
	`*design`	After risk assessment	Guide dev on test strategy	`docs/qa/assessments/{epic}.{story}-test-design-{YYYYMMDD}.md`
Development	`*trace`	Mid-implementation	Verify test coverage	`docs/qa/assessments/{epic}.{story}-trace-{YYYYMMDD}.md`
	`*nfr`	While building features	Catch quality issues early	`docs/qa/assessments/{epic}.{story}-nfr-{YYYYMMDD}.md`
Review	`*review`	Story marked complete	Full quality assessment	QA Results in story + gate file
Post-Review	`*gate`	After fixing issues	Update quality decision	Updated `docs/qa/gates/{epic}.{story}-{slug}.yml`

Example Commands

# Planning Stage - Run these BEFORE development starts
@qa *risk {draft-story}     # What could go wrong?
@qa *design {draft-story}   # What tests should we write?

# Development Stage - Run these DURING coding
@qa *trace {story}          # Are we testing everything?
@qa *nfr {story}            # Are we meeting quality standards?

# Review Stage - Run when development complete
@qa *review {story}         # Comprehensive assessment + refactoring

# Post-Review - Run after addressing issues
@qa *gate {story}           # Update gate status

Quality Standards Enforced

Quinn enforces these test quality principles:

No Flaky Tests: Ensures reliability through proper async handling
No Hard Waits: Dynamic waiting strategies only
Stateless & Parallel-Safe: Tests run independently
Self-Cleaning: Tests manage their own test data
Appropriate Test Levels: Unit for logic, integration for interactions, E2E for journeys
Explicit Assertions: Keep assertions in tests, not helpers

Gate Status Meanings

PASS: All critical requirements met, no blocking issues
CONCERNS: Non-critical issues found, team should review
FAIL: Critical issues that should be addressed (security risks, missing P0 tests)
WAIVED: Issues acknowledged but explicitly accepted by team

Special Situations

High-Risk Stories:

Always run *risk and *design before development starts
Consider mid-development *trace and *nfr checkpoints

Complex Integrations:

Run *trace during development to ensure all integration points tested
Follow up with *nfr to validate performance across integrations

Performance-Critical:

Run *nfr early and often during development
Don't wait until review to discover performance issues

Brownfield/Legacy Code:

Start with *risk to identify regression dangers
Use *review with extra focus on backward compatibility

Best Practices

Early Engagement: Run *design and *risk during story drafting
Risk-Based Focus: Let risk scores drive test prioritization
Iterative Improvement: Use QA feedback to improve future stories
Gate Transparency: Share gate decisions with the team
Continuous Learning: QA documents patterns for team knowledge sharing
Brownfield Care: Pay extra attention to regression risks in existing systems

Output Paths Reference

Quick reference for where Test Architect outputs are stored:

*risk-profile  → docs/qa/assessments/{epic}.{story}-risk-{YYYYMMDD}.md
*test-design   → docs/qa/assessments/{epic}.{story}-test-design-{YYYYMMDD}.md
*trace         → docs/qa/assessments/{epic}.{story}-trace-{YYYYMMDD}.md
*nfr-assess    → docs/qa/assessments/{epic}.{story}-nfr-{YYYYMMDD}.md
*review        → QA Results section in story + gate file reference
*gate          → docs/qa/gates/{epic}.{story}-{slug}.yml

Technical Preferences System

BMad includes a personalization system through the technical-preferences.md file located in .bmad-core/data/ - this can help bias the PM and Architect to recommend your preferences for design patterns, technology selection, or anything else you would like to put in here.

Using with Web Bundles

When creating custom web bundles or uploading to AI platforms, include your technical-preferences.md content to ensure agents have your preferences from the start of any conversation.

Core Configuration

The bmad-core/core-config.yaml file is a critical config that enables BMad to work seamlessly with differing project structures, more options will be made available in the future. Currently the most important is the devLoadAlwaysFiles list section in the yaml.

Developer Context Files

Define which files the dev agent should always load:

devLoadAlwaysFiles:
  - docs/architecture/coding-standards.md
  - docs/architecture/tech-stack.md
  - docs/architecture/project-structure.md

You will want to verify from sharding your architecture that these documents exist, that they are as lean as possible, and contain exactly the information you want your dev agent to ALWAYS load into its context. These are the rules the agent will follow.

As your project grows and the code starts to build consistent patterns, coding standards should be reduced to include only the standards the agent still needs enforced. The agent will look at surrounding code in files to infer the coding standards that are relevant to the current task.

Getting Help

Discord Community: Join Discord
GitHub Issues: Report bugs
Documentation: Browse docs
YouTube: BMadCode Channel

Conclusion

Remember: BMad is designed to enhance your development process, not replace your expertise. Use it as a powerful tool to accelerate your projects while maintaining control over design decisions and implementation details.

user-guide.md 22 KB Link permanente Histórico Em bruto