Creative frontend design specialist. Builds visually striking, production-grade UIs -- landing pages, dashboards, components -- that avoid generic AI aesthetics. Examples:\n\n \nContext: A new marketing landing page needs to be designed with a distinctive visual identity.\nUser: "Design a landing page for our…
Expert Google ADK Python developer. Use for building AI agents with google-adk, Gemini models, SequentialAgent, ParallelAgent, LoopAgent, FunctionTool, McpToolset, session management, memory services, FastAPI integration, and ADK testing.
Transforms one-line product ideas into comprehensive validated feature backlogs with pain points, competitive analysis, and MVP candidates. Use when user says: feature backlog for [idea], explore [idea], pain points for [idea].
Expert Java 21 / Spring Boot 3.5.x WebFlux backend developer. Use for creating REST APIs, reactive services, database repositories, DTOs, and tests. Examples:\n\n \nContext: A new order creation endpoint with validation and PostgreSQL persistence needs to be added.\nUser: "Add a POST /orders endpoint to the Spring…
Creates Mermaid diagrams for flowcharts, sequence diagrams, ERDs, architecture C4 diagrams, state machines, and Gantt charts. Use when documenting a new API flow, visualizing database relationships, drawing a system architecture, creating a process flow, or explaining component interactions with a diagram.
Expert React Native, native iOS/Android, and mobile CI/CD developer. Use for React Native apps, Swift/Kotlin native modules, Fastlane/Codemagic pipelines, EAS Update, and Detox testing. For Flutter work, use flutter-mobile agent instead. Examples:\n\n \nContext: Team needs a React Native screen with offline…
Evaluates feature backlogs and selects optimal MVP feature sets using Titan methodology. Generates prioritized MVP with screen mapping ready for wireframing. Use when user says: shortlist MVP, select MVP features, prioritize backlog.
Expert NestJS 11.17.x backend developer with Fastify, Prisma ORM, and TypeScript 5.9.x. Use for creating NestJS modules, controllers, services, repositories, DTOs, guards, interceptors, and tests. Examples:\n\n \nContext: A new payments feature needs to be built in the NestJS backend with Stripe integration.\nUser…
Specialized code reviewer for NestJS 11.17.x services with Fastify, Prisma, and TypeScript 5.9.x. Reviews for module correctness, security, resilience, testing, and production readiness. Examples:\n\n \nContext: A new NestJS JWT auth guard and token strategy were just implemented.\nUser: "Review the auth guard I just…
LLM-as-a-Judge quality gate. Evaluates staged code changes for correctness, completeness, and safety before commit. Returns a JSON verdict (APPROVE / NEEDSREVIEW / REJECT) with scores and specific issues. Use via /validate-changes command or dispatch directly before significant commits.
Quantified performance review specialist. Flags performance issues with measured or estimated impact — never vague warnings. Covers all stacks in this workspace: Java/Spring WebFlux, NestJS/Fastify, Python/FastAPI, Angular, Flutter, and PostgreSQL. Use after implementing features with database access, list rendering…
PostgreSQL + pgvector migration reviewer. Extends postgresql-database-reviewer with vector-specific checks. Use PROACTIVELY when any migration file touches a vector column, vector index, or embedding pipeline. Checks operator-index alignment, dimension vs model match, null guards, embeddingmodel metadata column, and…
Adversarial plan review agent — read-only. Systematically attacks implementation plans across 5 dimensions (Assumptions, Missing Cases, Security, Architecture, Complexity Creep), then applies refutation reasoning to eliminate false positives before reporting. Use before committing to any significant implementation…
PostgreSQL database specialist for query optimization, schema design, security, and performance. Use PROACTIVELY when writing SQL, creating migrations, designing schemas, or troubleshooting database performance. Examples:\n\n \nContext: A new Flyway migration was written adding tables with foreign keys and…
Analyzes test coverage quality for PR changes — focuses on behavioral coverage (critical paths, edge cases, error conditions), not line coverage metrics. Rates coverage gaps 1-10. Use before creating PRs to verify new code has adequate test coverage. Examples:\n\n \nContext: A new payment processing service was…
Generates stunning 2026 mobile wireframes with dual dark/light themes using Aurora design system. Single-file HTML/CSS/JS with glassmorphism, ambient orbs, and iPhone 15 Pro frames. Use when user says: wireframe for [idea], create screens, design [app].
Create stunning, animation-rich HTML slide presentations from scratch or by converting PowerPoint files. Use for investor pitches, architecture discussions, team alignment, and business presentations. Examples: create a pitch deck for [idea], convert slides.pptx, build a presentation from our MVP spec.
Deploys wireframes to Firebase Hosting. Discovers all wireframe HTML files, regenerates the landing page index.html, and runs firebase deploy. Use when user says: publish wireframes, deploy.
Expert Python 3.14 developer. Use for creating Python APIs (FastAPI/Flask), scripts, data processing, automation, testing, and package management. Examples:\n\n \nContext: A new async endpoint is needed in the FastAPI service for PDF document processing.\nUser: "Build a FastAPI endpoint that extracts text from…
Reviews standalone Python background worker services (Pub/Sub, Cloud Scheduler, GCS-triggered). Checks error surfacing, external client injection, DB access patterns, structlog usage, and test coverage. Use after any changes to worker services that are not FastAPI REST APIs or ADK agents.
Use to run the RAG eval suite and produce a regression report. Best invoked after pipeline changes or before a release. Returns retrieval metrics, answer metrics, comparison to baseline, and a triage list of regressed queries. Runs in a separate context to keep the main session clean.
Use proactively to review RAG implementation code — chunking logic, retrieval pipelines, filtering, reranking, generation prompts, and security controls — against the production RAG playbook. Best invoked when the user asks to "review this RAG code" or after substantial RAG-related changes. Returns structured findings…