@memhq/llamaindex

MemHQ adapter for LlamaIndex.js. Drop-in chat memory and retriever backed by MemHQ.

Install

npm install @memhq/llamaindex @memhq/sdk llamaindex

Quickstart — Chat memory

import { OpenAIAgent } from "llamaindex";
import { MemHQChatMemory } from "@memhq/llamaindex";

const memory = new MemHQChatMemory({
  apiKey: process.env.MEMHQ_API_KEY!,
  sessionId: "conv_42",
  userId: "user_42",
});

const agent = new OpenAIAgent({
  tools: [],
  memory,
});

const reply = await agent.chat({ message: "What's my favorite coffee?" });

Quickstart — Retriever

import { RetrieverQueryEngine } from "llamaindex";
import { MemHQRetriever } from "@memhq/llamaindex";

const retriever = new MemHQRetriever({
  apiKey: process.env.MEMHQ_API_KEY!,
  userId: "user_42",
  limit: 8,
});

const engine = new RetrieverQueryEngine(retriever);
const answer = await engine.query({ query: "Summarize pricing discussion" });

Configuration

| Option | Description | | --- | --- | | apiKey | MemHQ project API key. Falls back to MEMHQ_API_KEY. | | userId | External user id. Required. | | sessionId | Thread anchor for the memory adapter. | | recallLimit | Memories surfaced per memory read. Default 10. | | limit | Memories returned per retriever query. Default 10. | | mode | "hybrid" (default), "vector", or "lexical". | | baseUrl | Override the API base. |

Reference

Full reference: https://docs.memhq.ai/sdks/llamaindex-js