bmo-agent-setup

Code quality review focused on naming, complexity, dead code, test quality, and project conventions. Complements staff-engineer review. Use when the user wants a dedicated quality pass on implementation changes — phrases like "code quality review", "check code quality", "review naming and complexity", "quality pass", "check conventions", "dead code check", or "review test quality". Independent of the dev-team workflow; can be invoked directly on any bmo issue.

Orchestrate a software development agent team: @staff-engineer (TDDs + code review), @project-manager (bmo planning), @ux-designer (UX specs), @senior-engineer (implementation), @qa-engineer (testing). Use whenever the user wants to plan AND execute work — feature development, migrations, refactors, bug batches, or any multi-issue project. Trigger on: "use the agent team", "plan and execute", "have the team work on", "spin up engineers", "run the dev team", "parallel development", "multi-agent execution", or when the user describes work that clearly needs both planning and execution.

Distributed systems correctness analysis: safety/liveness properties, consistency models, fault tolerance, and TLA+ formal verification. Use when work touches distributed coordination, consensus, replication, shared mutable state across nodes, consistency model choices, CRDTs, distributed transactions, or ordering/causality. Trigger on: "check for distributed systems issues", "review DS correctness", "write a TLA+ spec", "analyze consensus", "review for safety violations", "distributed systems review", or when implementation touches concurrent state across nodes or processes.

Bootstrap project specification files in docs/spec/ by spawning staff-engineer agents in parallel. Use when the user wants to initialize, generate, or bootstrap project specs — phrases like "document the project", "initialize specs", "generate specs", "create project specifications", "bootstrap docs/spec", "populate specs", or "set up project documentation".

Orchestrate a software development agent team consisting of @staff-engineer (design + review), @project-manager (planning), @ux-designer (UX design), @senior-engineer (implementation), and @qa-engineer (testing). Use this skill whenever the user wants to plan AND execute a body of work using the agent team pattern — including feature development, migrations, refactors, bug fix batches, or any multi-issue project. Use this skill IF the user request benefits from research, planning, multiple indepedent tasks, and validation. Trigger on phrases like "use the agent team", "plan and execute", "have the team work on", "spin up engineers", "run the dev team on this", or when the user describes work that clearly needs both planning and execution. Also trigger when the user references @project-manager and @senior-engineer together, or asks for "parallel development", "multi-agent execution", or "agent swarm".

Bootstrap the project specification files in docs/spec/ by spawning 5 @staff-engineer agents in parallel. Use this skill when the user wants to initialize, generate, or bootstrap project specs — including phrases like "document the project", "initialize specs", "generate specs", "create project specifications", "bootstrap docs/spec", "populate specs", or "set up project documentation".