Support prompt configuration in createPrompt (e.g. model, template, output schema)

[moretti]

Checked

• I searched existing ideas and did not find a similar one
• I added a very descriptive title
• I've clearly described the feature request and motivation for it

Feature request

Currently the createPrompt method in the LangChain JavaScript SDK only allows creating prompts by name, without any additional configuration such as model, template, etc...
It would be helpful if createPrompt accepted additional parameters to define a complete prompt in a single step.

Motivation

When working with LangSmith programmatically, it’s frustrating to only be able to create a barebones prompt by name and then have to manually update its configuration afterward—especially for the output schema, which is often generated directly from the codebase (e.g a Zod object converted to JSON Schema).

Proposal (If applicable)

import { Client } from 'langsmith';

const client = new Client();
await client.createPrompt('extract-contact-info', {
  description: 'Extracts name and email from a block of text.',
  model: new ChatOpenAI({ modelName: "gpt-4o-mini" }),
  template: ChatPromptTemplate.fromMessages([
    ["system", "You extract structured contact info from user input."],
    ["user", "Extract name and email from this text:\n\n{input}"],
  ]),
  outputSchema: {
    type: "object",
    properties: {
      name: { type: "string" },
      email: { type: "string", format: "email" }
    },
    required: ["name", "email"]
  },
});