Architect and review production AI applications with LLM integrations and RAG pipelines.
npx skills add https://github.com/iuliandita/skills --skill ai-ml-iuliandita
Please help me install this Agent Skill. Skill: ai-ml Source: https://github.com/iuliandita/skills/tree/main/skills/ai-ml Command: npx skills add https://github.com/iuliandita/skills --skill ai-ml-iuliandita
This skill addresses the complexity of building production-grade AI applications by providing a structured framework for model integration, RAG pipelines, and agentic workflows, preventing common pitfalls like hallucinations and inefficient token usage.
Use the ai-ml skill to review my current RAG pipeline architecture and suggest improvements for retrieval accuracy.
High-intent search queries and answers about installing and using this skill.
Architecting a production RAG pipeline requires structured chunking, embedding, and retrieval practices to ensure accurate retrieval and cost-effective performance. Proper abstraction selection prevents common pitfalls like inefficient token usage and hallucinations.
Evaluating LLM outputs against safety guardrails involves validating structured output and using built-in checklists to manage token budgets. This rigorous validation prevents prompt injection and ensures reliable, production-grade model performance.
Selecting the right abstraction level for multi-agent systems ranges from using raw provider SDKs to stateful agent frameworks like LangGraph. Choosing correctly ensures reliable agent loop management and robust workflow execution.
Provider-specific SDK patterns are required for local inference setup and model routing in production AI applications. Adhering to these patterns ensures structured integration and optimal routing across different model providers.
LLM integrations suffer from hallucinations and inefficient token usage due to poorly structured model integration and agentic workflows. Implementing proper RAG hygiene and agent loop management prevents these common AI application pitfalls.
A stateful agent framework like LangGraph is needed for production AI architecture when managing complex agentic workflows and stateful loops. Raw provider SDKs may suffice for simpler LLM integration tasks without state management.
adelproDefine standardized MCP server designs specifying transports, resources, error handling, naming conventions, input schemas, outputs, and testing/security practices.
DariusCorvusDetect deployment platforms and write CLAUDE.md Deploy Configuration for gstack projects.
bcastelinoDesign and optimize RAG systems for LLM applications.
hainamchungDesign and implement production-grade LLM applications, RAG systems, and intelligent agents.
