bmad-cowork

Run a BMAD implementation sprint autonomously — spawns the full agent team as Cowork teammates and orchestrates phase-by-phase execution with handoffs.

Initialize a BMAD Cowork project — creates the project structure, spawns the agent team, and begins the first phase. Use this to start any new BMAD project from scratch.

Show current BMAD project status — active phase, completed artifacts, team activity, and next steps.

Use this agent when the user needs a quick fix, simple task, one-off script, or any work that doesn't need full BMAD ceremony. <example> User: "Quick fix — the login button color is wrong" Action: barry-quick locates the issue and fixes it immediately without story ceremony. </example> <example> User: "Write a quick script to migrate the config format" Action: barry-quick specs, implements, and tests the migration script in one pass. </example> <example> User: "Skip the ceremony, just add input validation to the signup form" Action: barry-quick writes a brief spec, implements the validation, and verifies it works. </example>

Use this agent when the user needs sprint planning, story preparation, story file creation, or to advance to the next story in development. <example> User: "Plan the sprint and create stories for Epic 1" Action: bob-sm reads epics and architecture, then creates detailed story files with extracted context. </example> <example> User: "Prepare the next story for development" Action: bob-sm identifies the next unpicked story and prepares it with full architecture context. </example> <example> User: "Create stories for the authentication epic" Action: bob-sm breaks down the epic into sequenced stories with tasks and acceptance criteria. </example>

Use this agent when the user needs story implementation, feature development, coding tasks, or to start development on a prepared story. <example> User: "Implement story 1.1" Action: james-dev reads the story file and implements each task, updating the story checkboxes as he goes. </example> <example> User: "Start development on the authentication story" Action: james-dev reads the prepared story, implements code task by task with tests, and marks progress. </example> <example> User: "Code the user registration endpoint" Action: james-dev implements the endpoint following architecture specs from the story file. </example>

Use this agent when the user needs PRD creation, product requirements definition, epic planning, product validation, or product strategy. <example> User: "Create a PRD for the fitness tracking app" Action: john-pm reads the project brief and produces a comprehensive PRD with epics and user stories. </example> <example> User: "Define epics and requirements for our authentication system" Action: john-pm creates structured requirements organized by epic with acceptance criteria. </example> <example> User: "Validate the PRD against the project brief" Action: john-pm runs the PRD quality checklist and identifies gaps or misalignments. </example>

Use this agent when the user needs brainstorming, market research, domain analysis, competitive analysis, or project brief creation. <example> User: "Can you brainstorm ideas for a fitness tracking app?" Action: mary-analyst conducts structured brainstorming with market context and outputs a project brief. </example> <example> User: "Research the competitor landscape for project management tools" Action: mary-analyst performs competitive analysis and produces docs/competitor-analysis.md. </example> <example> User: "Create a project brief for our new e-commerce platform" Action: mary-analyst gathers requirements through elicitation and produces docs/brief.md. </example>

Use this agent when the user needs technical documentation, README files, API docs, user guides, or architecture decision records. <example> User: "Write the README for the project" Action: paige-writer reads the codebase and produces a comprehensive README.md. </example> <example> User: "Document the API endpoints" Action: paige-writer reads the API specification and code to produce API documentation. </example> <example> User: "Create an architecture decision record for choosing PostgreSQL" Action: paige-writer creates a structured ADR in docs/adr/ with context, decision, and consequences. </example>

Use this agent when the user needs QA review, story validation, quality checks, test verification, or gate decisions on completed work. <example> User: "Review story 1.1" Action: quinn-qa reads the story and implementation, validates against acceptance criteria, and provides a gate decision. </example> <example> User: "QA check the authentication implementation" Action: quinn-qa runs tests, reviews code quality, and checks acceptance criteria compliance. </example> <example> User: "Gate check — is this story ready for production?" Action: quinn-qa performs a comprehensive quality review and issues PASS, CONCERNS, FAIL, or WAIVED. </example>

Use this agent when the user needs UX design, wireframes, frontend specifications, UI design, or user flow diagrams. <example> User: "Design the UX for the onboarding flow" Action: sally-ux creates wireframes and user flow diagrams for the onboarding experience. </example> <example> User: "Create a frontend spec for the dashboard" Action: sally-ux produces a detailed frontend specification with component hierarchy and interactions. </example> <example> User: "Generate UI prompts for v0 to build the landing page" Action: sally-ux creates AI UI generation prompts optimized for v0/Lovable/Bolt. </example>

Use this agent when the user needs system architecture design, technical architecture, backend architecture, tech stack decisions, or infrastructure planning. <example> User: "Create the architecture for the fitness tracking app" Action: winston-architect reads the PRD and produces a complete sharded architecture document set. </example> <example> User: "Design the backend architecture and API specification" Action: winston-architect creates backend-architecture.md and api-specification.md with full details. </example> <example> User: "What tech stack should we use for this project?" Action: winston-architect evaluates options and documents decisions in tech-stack.md with rationale. </example>

This skill activates when the user asks about "architecture extraction", "story context", "dev notes", "source citations", "context bridge", "relevant architecture", or needs guidance on how BMAD extracts and embeds relevant architecture context into story files for developers.

This skill activates when the user discusses BMAD methodology, project phases, agent coordination, phase gates, workflow routing, or asks about "which BMAD agent", "what phase am I in", "BMAD workflow", "phase transition", "agent handoff". Provides knowledge about BMAD's 4-phase methodology and 9 specialized agents.

This skill activates when the user asks about "story creation", "sprint planning", "story development cycle", "story validation", "code review workflow", "dev story", "create next story", "story template", or needs guidance on BMAD's implementation workflow patterns.