Llm Architect

What it does

The omer-metin-llm-architect skill assists in designing and implementing Large Language Model (LLM) architectures. It helps users define the components of an LLM system, including model selection, prompt engineering strategies, and integration with external tools or data sources. This skill facilitates building more complex AI agent workflows that leverage LLMs effectively.

When to use it

• Designing a new AI Agent: Use this when you're starting from scratch and need help structuring the LLM components of your agent.
• Optimizing Existing Workflows: If an existing workflow is underperforming, this skill can assist in re-architecting the LLM usage for better results.
• Integrating External Data: When needing to connect an LLM with specific databases or APIs, this skill helps define the integration points and data flow.
• Experimenting with Models: To quickly prototype different LLM configurations and compare their performance on a given task.

Key capabilities

• Model Selection Guidance
• Prompt Engineering Strategy Definition
• Integration Design (External Tools/Data)
• LLM System Architecture Planning

Example prompts

• "I need to build an agent that summarizes legal documents. Suggest an LLM architecture."
• "How can I integrate a vector database with my LLM workflow for semantic search?"
• "What are some prompt engineering techniques I could use to improve the accuracy of my chatbot's responses?"

Tips & gotchas

The skill’s effectiveness depends on providing clear context about the desired agent functionality and any existing constraints. It is best used in conjunction with other skills focused on specific LLM tasks (e.g., prompt generation, data extraction).