Applies backend engineering traits for APIs, data modeling, reliability, security, and observability. Use when building or reviewing routes, services, schemas, migrations, queues, concurrency, or when the user asks for a backend engineer perspective.
Applies frontend engineering traits for UI quality, accessibility, perceived performance, and state scope. Use when building or reviewing components, pages, styling, client-side data fetching, Core Web Vitals, interaction states, or when the user asks for a frontend engineer perspective.
Applies AI/ML engineering traits for evaluation rigor, data quality, model selection, and production ML/LLM systems. Use when building or reviewing ML pipelines, RAG, embeddings, prompts, model choice, offline/online evals, inference cost or latency, hallucination risk, or when the user asks for an AI/ML engineer perspective.
Applies DevOps and platform engineering traits for automation, reliability, security, cost, and developer experience. Use when working on CI/CD, infrastructure as code, deployments, observability, incidents, runbooks, secrets, cloud cost, or when the user asks for a DevOps or platform engineer perspective.
Applies fullstack engineering traits for end-to-end design, seam placement, and cross-boundary debugging. Use when reviewing or designing features across client and server, deciding where logic lives (client, server, edge, background job), tracing bugs through the full stack, or when the user asks for a fullstack engineer perspective.
Perform a structured code review of changes, checking for correctness, style, tests, and potential issues.
Read the user's coding preferences from /memory/coding-prefs.md before making non-trivial style decisions, and append new preferences when the user gives durable feedback.
Dispatch a batch of tasks to subagents in parallel with bounded concurrency. Returns a summary object with {total, completed, failed, results[]} — iterate `.results` for per-task output.