LangChain.js Integration

The @dakera-ai/langchain package provides LangChain.js components backed by Dakera — persistent agent memory and server-side vector search for TypeScript applications.

Quick Start

npm install @dakera-ai/langchain

Requires Node.js ≥ 18 and a running Dakera server.

import { DakeraMemory, DakeraVectorStore } from "@dakera-ai/langchain";

const memory = new DakeraMemory({
  apiUrl: "http://localhost:3300",
  apiKey: "dk-mykey",
  agentId: "my-agent",
});

const vectorstore = new DakeraVectorStore({
  apiUrl: "http://localhost:3300",
  apiKey: "dk-mykey",
  namespace: "my-docs",
});

Features

DakeraMemory — extends BaseMemory with semantic recall
DakeraVectorStore — extends VectorStore for RAG (server-side embeddings)
DakeraSessionManager — conversation session lifecycle
DakeraKnowledgeGraph — entity-relationship graph operations
DakeraEntityExtractor — automatic entity extraction from text
DakeraNamespaceManager — multi-tenant namespace isolation
DakeraAgentTools — LangChain tools wrapping Dakera operations

Examples

Conversation chain with persistent memory

import { ChatOpenAI } from "@langchain/openai";
import { ConversationChain } from "langchain/chains";
import { DakeraMemory } from "@dakera-ai/langchain";

const memory = new DakeraMemory({
  apiUrl: "http://localhost:3300",
  apiKey: "dk-mykey",
  agentId: "chat-agent",
  recallK: 5,
  importance: 0.7,
});

const chain = new ConversationChain({
  llm: new ChatOpenAI({ model: "gpt-4o" }),
  memory,
});

const res = await chain.call({ input: "My name is Alice." });

// Later — memory persists across restarts
const res2 = await chain.call({ input: "What's my name?" });

RAG retriever

import { DakeraVectorStore } from "@dakera-ai/langchain";

const vectorstore = new DakeraVectorStore({
  apiUrl: "http://localhost:3300",
  apiKey: "dk-mykey",
  namespace: "my-docs",
});

// Add documents
await vectorstore.addDocuments([
  { pageContent: "Dakera handles embeddings server-side.", metadata: {} },
  { pageContent: "Memory decay removes stale data automatically.", metadata: {} },
]);

// Retrieve
const retriever = vectorstore.asRetriever({ k: 4 });
const docs = await retriever.getRelevantDocuments("How does Dakera embed?");

Session management

import { DakeraSessionManager } from "@dakera-ai/langchain";

const sessions = new DakeraSessionManager({
  apiUrl: "http://localhost:3300",
  apiKey: "dk-mykey",
  agentId: "my-agent",
});

const session = await sessions.start({ metadata: { user: "alice" } });
// ... use session.id with memory operations ...
await sessions.end(session.id, "Completed onboarding flow");

API Reference

DakeraMemory options

Parameter	Type	Default	Description
`apiUrl`	string	—	Dakera server URL
`apiKey`	string	`""`	API key
`agentId`	string	—	Agent identifier for memory namespacing
`recallK`	number	`5`	Memories to surface per turn
`importance`	number	`0.7`	Importance assigned to stored memories
`memoryKey`	string	`"history"`	Key injected into the prompt

DakeraVectorStore options

Parameter	Type	Default	Description
`apiUrl`	string	—	Dakera server URL
`apiKey`	string	`""`	API key
`namespace`	string	—	Vector namespace to read/write

Configuration

DAKERA_API_URL=http://localhost:3300
DAKERA_API_KEY=dk-your-key

Set these in your .env file. The client reads them when constructor options are not passed explicitly.