LangChaingo API (JS SDK)

BosBase exposes the /api/langchaingo endpoints so you can run LangChainGo powered workflows without leaving the platform. The JS SDK wraps these endpoints with the pb.langchaingo service.

The service exposes two high-level methods:

Both methods accept an optional model block:

interface LangChaingoModelConfig {
    provider?: "openai" | "ollama" | string;
    model?: string;
    apiKey?: string;
    baseUrl?: string;
}

If you omit the model section, BosBase defaults to provider: "openai" and model: "gpt-4o-mini" with credentials read from the server environment. Passing an apiKey lets you override server defaults on a per-request basis.

Text + Chat Completions

import Client from "bosbase";

const pb = new Client("http://localhost:8090");

const completion = await pb.langchaingo.completions({
    model: { provider: "openai", model: "gpt-4o-mini" },
    messages: [
        { role: "system", content: "Answer in one sentence." },
        { role: "user", content: "Explain Rayleigh scattering." },
    ],
    temperature: 0.2,
});

console.log(completion.content);

The completion response mirrors the LangChainGo ContentResponse shape, so you can inspect the functionCall, toolCalls, or generationInfo fields when you need more than plain text.

Retrieval-Augmented Generation (RAG)

Pair the LangChaingo endpoints with the /api/llm-documents store to build RAG workflows. The backend automatically uses the chromem-go collection configured for the target LLM collection.

const answer = await pb.langchaingo.rag({
    collection: "knowledge-base",
    question: "Why is the sky blue?",
    topK: 4,
    returnSources: true,
    filters: {
        where: { topic: "physics" },
    },
});

console.log(answer.answer);
for (const source of answer.sources ?? []) {
    console.log(source.score?.toFixed(3), source.metadata?.title);
}

Set promptTemplate when you want to control how the retrieved context is stuffed into the answer prompt:

await pb.langchaingo.rag({
    collection: "knowledge-base",
    question: "Summarize the explanation below in 2 sentences.",
    promptTemplate: `Context:\n{{.context}}\n\nQuestion: {{.question}}\nSummary:`,
});

Dart SDK

Looking for the Dart equivalent? See dart-sdk/doc/LANGCHAINGO_API.md for identical request/response shapes and Flutter-ready examples.