What problem does it solve? Building your first LLM agents involves many small decisions—persona definition, run versus stream execution, sampling parameters, and context injection—that are easy to get wrong without guidance. This Skill provides working Agno code patterns for each of these foundational choices. ## Core Features & Use Cases - Agent Persona Setup: Define an Agent with name, description, and instructions to separate who the agent is from how it behaves. - Run vs Stream Execution: Choose between full-response runs and token-by-token streaming for better interactive UX. - Sampling Parameter Control: Configure temperature, top_p, top_k, and max_tokens on models like OpenAIChat to balance determinism, creativity, and cost. - Context Engineering: Inject static context dictionaries and datetime into agents while keeping instructions short and specific. - Use Case: A developer learning the Agno framework can compare temperature=0 versus temperature=1 outputs and toggle streaming to understand how each parameter changes agent behavior. ## Quick Start Create a basic Agno agent with a mentor persona, streaming output, and tuned sampling parameters using the patterns in this skill.
