LLM & Agents

Perform systematic code review following Phoenix/FAAD HMI standards and best practices. Use when reviewing pull requests, checking code quality, or validating implementations before commit. Checks MobX patterns, UI components, architecture, testing, and common pitfalls.

AI SDK v6 development patterns and guidance. Triggers on AI SDK, generateText, streamText, ToolLoopAgent, tool, agent, LLM, structured output, generateObject, useChat, @ai-sdk/react, createAgentUIStreamResponse, Output.object, stepCountIs, UIMessage, ModelMessage keywords.

Create interactive algebra tiles manipulatives using p5.js where students drag tiles and see cancellations when opposites overlap.

Analyzes property listings to extract key data, generate comparisons, identify insights, and create professional real estate documentation for agents and investors.

An important skill, to help agents assist doctors with diagnosising patients.

Scaffolds a new Python tool utility in 'src/tools/' for agents to call. Tools are utility functions (web scraping, PDF parsing, database queries) that agents use. Use when creating utilities that agents depend on.

Pattern for spawning parallel subagents efficiently. Use when you need multiple independent tasks done concurrently.

Expert knowledge of ZIP-based archive synchronization system. Use when working with archive agents, manifests, sync operations, or troubleshooting archive-related issues.

Evaluates and improves Claude Skills across all ~/.claude/ locations

Opens a named terminal tab in VSCode via Antigravity Bridge (port 4000). Use when spawning agent terminals for swarm workers.

Master RL theory - MDPs, value functions, Bellman equations, value/policy iteration, TD

Use when executing implementation plans with independent tasks in the current session - dispatches fresh subagent for each task with code review between tasks, enabling fast iteration with quality gates

Virtual test environment for Claude Code hooks, plugins, and agents. Use when testing or validating plugin components without affecting production.

Analyze Claude Code plugins for quality, DX, and architecture. Use when "review plugin quality", "audit plugins", "analyze the marketplace", "check plugins against Anthropic standards", "improve plugin design", or evaluating plugin architecture. Provides systematic analysis methodology with validation framework.

Use when working with unit tests in any capacity - reading, writing, updating, planning, evaluating, or reviewing tests - applies TDD principles with focused test coverage, filtering out redundant tests while ensuring critical paths, edge cases, and state changes are verified

Manages ticket CRUD operations for the project management system. Handles creating, reading, updating, listing, and searching tickets stored in .claude/data/tickets/.

Production-grade agentic system development with LlamaIndex in Python. Covers semantic ingestion (SemanticSplitterNodeParser, CodeSplitter, IngestionPipeline), retrieval strategies (BM25Retriever, hybrid search, alpha weighting), PropertyGraphIndex with graph stores (Neo4j), context RAG (RouterQueryEngine, SubQuestionQueryEngine, LLMRerank), agentic orchestration (ReAct, Workflows, FunctionTool), and observability (Arize Phoenix). Use when asked to "build a LlamaIndex agent", "set up semantic chunking", "index source code", "implement hybrid search", "create a knowledge graph with LlamaIndex", "implement query routing", "debug RAG pipeline", "add Phoenix observability", or "create an event-driven workflow". Triggers on "PropertyGraphIndex", "SemanticSplitterNodeParser", "CodeSplitter", "BM25Retriever", "hybrid search", "ReAct agent", "Workflow pattern", "LLMRerank", "Text-to-Cypher".

Use this skill when the user asks to create a new skill, custom command, or wants to know how skills work in Claude Code.

Publishing assistant for Thomas's Astro-based GitHub Pages site (readme-pages). Creates publish-ready Markdown files for notes, recipes, and pages. Use when asked to write blog posts, articles, notes, recipes, or content for the readme-pages site at thomasrohde.github.io/readme-pages. Also use for content requests mentioning "publish", "blog", or "my site".

Generate professional prompts for Nano Banana Pro image generation. Use when the user wants to create images, thumbnails, infographics, edit photos, generate character-consistent content, create storyboards, or any visual asset using Nano Banana Pro. Transforms vague requests into detailed, effective prompts following Nano Banana Pro's "thinking model" best practices.