Prompt Templates ​

Hopfield always uses string literal types for prompt templates, so developers have visibility into the tranformations performed on their inputs.

Usage ​

Check out how our types look when you use a template:

ts
import hop from "hopfield";
import openai from "hopfield/openai";
 
const template = hop.client(openai).template();
 
const description = template.enum("The category of the message.");
          
const description: "The category of the message. This must always be a possible value from the `enum` array."

ts
import hop from "hopfield";
import openai from "hopfield/openai";
 
const template = hop.client(openai).template();
 
const description = template.enum("The category of the message.");
          
const description: "The category of the message. This must always be a possible value from the `enum` array."

You can see above that the description has type hints to tell you exactly what the transformation was. In this case, the template appended This must always be a possible value from the enum array. to the input string.

This template is usually used with complex function calls. See the next section for more information.

Composability ​

We will be building on top of the Prompt Templating primitive with features which have more complex transformations. Specifically, we will be shipping best practices for few-shot prompting and RAG, so chat messages are strongly typed and adhere to emerging industry standards.