Integrations

npm install novyx

import { Novyx } from "novyx";

const nx = new Novyx({ apiKey: "nram_your_key" });

await nx.remember({ observation: "User prefers dark mode", tags: ["prefs"] });

const results = await nx.recall("user preferences");

• • Full TypeScript support with complete type definitions
• • Zero dependencies
• • 50+ methods — full API parity with Python SDK (eval methods coming soon)
• • Dual CJS/ESM — works in Node.js 18+ and browsers

pip install novyx

novyx remember "User prefers dark mode" --tags ui
novyx recall "user preferences"
novyx rollback "2 hours ago"
novyx spaces list
novyx audit --last 10

• • 24+ commands — memories, spaces, rollback, traces, audit, and more
• • Interactive and scriptable
• • Same package as Python SDK (pip install novyx)

Agent framework with built-in persistent memory — auto-recall, auto-remember, @tool decorator, multi-provider LLM support (OpenAI, Anthropic, LiteLLM).

pip install novyx-agent

• • Tiered recall — memories ranked [strong], [moderate], [background] with background capped at 2
• • Auto-remember — agent stores important context after every response
• • @tool decorator for custom tool functions
• • Multi-provider: OpenAI, Anthropic, and any LiteLLM-compatible model

pip install novyx-langchain

from novyx_langchain import NovyxMemory
from langchain.chains import ConversationChain
from langchain_openai import ChatOpenAI

memory = NovyxMemory(api_key="nram_your_key", session_id="user-123")
chain = ConversationChain(llm=ChatOpenAI(), memory=memory)
response = chain.invoke({"input": "Hello!"})

• • NovyxMemory — persistent semantic memory across conversations
• • Checkpointer — LangGraph stateful workflow integration
• • TracedMemory — RSA-signed execution traces
• • IntegrityMemory — tamper-proof audit trail

pip install novyx-crewai

from novyx_crewai import NovyxStorage
from crewai import Crew

storage = NovyxStorage(api_key="nram_your_key", memory_type="short_term")
crew = Crew(agents=[...], tasks=[...], memory=True, storage=storage)

• • Works with all 3 CrewAI memory types (short-term, long-term, entity)
• • Semantic search across agent memories
• • Tag-based filtering for scoped recall

pip install novyx-llamaindex

from novyx_llamaindex import NovyxChatStore
from llama_index.core.memory import ChatMemoryBuffer

chat_store = NovyxChatStore(api_key="nram_your_key")
memory = ChatMemoryBuffer.from_defaults(
    chat_store=chat_store,
    chat_store_key="user-123",
    token_limit=3000,
)

• • Drop-in persistent chat store for any LlamaIndex chat engine
• • Semantic recall via NovyxMemory (BaseMemory)
• • RAG retriever for memory-augmented pipelines
• • Rollback, audit trail, and cross-session persistence

pip install novyx-hygiene
hygiene install

• • Auto-saves before /compact via PreCompact hook
• • Auto-injects context on SessionStart (compact, clear, resume)
• • Context health scoring (freshness, file sprawl, mixed concerns)
• • Optional cloud sync with Novyx API key
• • Writes .claude/hygiene.md so Claude reads context automatically

OpenClaw

OpenClaw — open-source AI agent framework

Novyx ships 3 OpenClaw skills with persistent memory, rollback, and audit built in. Install and your agent gets long-term memory with zero configuration.

git clone https://github.com/novyxlabs/novyx-memory-skill.git extensions/novyx-memory
cd extensions/novyx-memory && npm install

{
  "extensions": {
    "novyx-memory": {
      "apiKey": "nram_your_key_here"
    }
  }
}

Templates

Ready-to-run examples to get started fast.