SmartFAQs.ai
Back to Learn
Intermediate

Dynamic Prompting

The programmatic construction of LLM instructions at runtime where prompt templates are injected with real-time context, retrieved documents, or user-specific state to adapt the model's behavior to specific execution paths. It allows an AI agent to shift its persona, constraints, or available data based on the logic flow of the pipeline.

Definition

The programmatic construction of LLM instructions at runtime where prompt templates are injected with real-time context, retrieved documents, or user-specific state to adapt the model's behavior to specific execution paths. It allows an AI agent to shift its persona, constraints, or available data based on the logic flow of the pipeline.

Disambiguation

Not static prompt engineering; it involves variable injection and conditional logic during code execution.

Visual Metaphor

"A digital billboard that changes its language, imagery, and product offers instantly based on the specific profile of the person standing in front of it."

Key Tools
LangChain PromptTemplatesLlamaIndexJinja2Semantic KernelHaystack
Related Connections

Conceptual Overview

The programmatic construction of LLM instructions at runtime where prompt templates are injected with real-time context, retrieved documents, or user-specific state to adapt the model's behavior to specific execution paths. It allows an AI agent to shift its persona, constraints, or available data based on the logic flow of the pipeline.

Disambiguation

Not static prompt engineering; it involves variable injection and conditional logic during code execution.

Visual Analog

A digital billboard that changes its language, imagery, and product offers instantly based on the specific profile of the person standing in front of it.

Related Articles