llm-application-dev

Plugin llm-application-dev from marketplace claude-code-workflows in shotcowboystyle/ag-skills. Commands read from the repo's manifest — 11 days ago.

Skills (8)

• embedding-strategies

  Choose an embedding model and a chunking strategy for semantic search and RAG. Use when choosing embedding models, splitting documents into chunks, reducing embedding dimensions, or evaluating embedding quality on a specific domain.

• hybrid-search-implementation

  Combine vector and keyword (BM25 / full-text) search with Reciprocal Rank Fusion, weighted fusion and reranking. Use when implementing RAG retrieval, when pure vector search misses exact terms (names, codes, IDs), or when tuning how ranked lists are fused.

• langchain-architecture

  Design LLM applications using LangChain 1.x and LangGraph for agents, memory, and tool integration. Use when building LangChain applications, implementing AI agents, or creating complex LLM workflows.

• llm-evaluation

  Implement comprehensive evaluation strategies for LLM applications using automated metrics, human feedback, and benchmarking. Use when testing LLM performance, measuring AI application quality, or establishing evaluation frameworks.

• prompt-engineering-patterns

  This skill should be used when the user asks to "optimize a prompt", "improve prompt performance", "design a prompt template", "write better prompts", "debug prompt issues", "use chain-of-thought", "structured prompting", "few-shot prompting", or wants to apply advanced prompt engineering patterns f

• rag-implementation

  Build Retrieval-Augmented Generation (RAG) systems for LLM applications with vector databases and semantic search. Use when implementing knowledge-grounded AI, building document Q&A systems, or integrating LLMs with external knowledge bases.

• similarity-search-patterns

  Implement nearest-neighbor search against a vector database (Pinecone, Qdrant, pgvector, Weaviate) and choose a distance metric. Use when building semantic search, writing vector store client code, or adding metadata filters to vector queries.

• vector-index-tuning

  Optimize vector index performance for latency, recall, and memory. Use when tuning HNSW parameters, selecting quantization strategies, or scaling vector search infrastructure.

Slash commands (3)

• /llm-application-dev:ai-assistant

  Build AI assistant application with NLU, dialog management, and integrations

• /llm-application-dev:langchain-agent

  Create LangGraph-based agent with modern patterns

• /llm-application-dev:prompt-optimize

  Optimize prompts for production with CoT, few-shot, and constitutional AI patterns

Subagents (3)

• ai-engineer

  Turns a feature requirement into an architecture decision for an LLM feature — plain call, chain, RAG or agent; which model tier; what to evaluate — and hands back a design the caller can implement. Use when planning a new LLM-powered feature or reworking one that is too slow, too costly or giving w

• prompt-engineer

  Write or revise one production prompt (system prompt, task prompt or template) for a stated task and target model, and return the full prompt text with test cases. Use when building an LLM feature that needs a new prompt, or when an existing prompt gives wrong, inconsistent or unparseable output.

• vector-database-engineer

  Design or review the vector search setup for one retrieval workload (embedding model, chunking, store, index parameters, hybrid search) and return a configuration recommendation with a measurement plan. Use when choosing or auditing a vector store configuration for RAG or semantic search.

Similar plugins