Skip to content

langchain-foxnose

LangChain integration for FoxNose — the serverless knowledge platform purpose-built as the knowledge layer for RAG and AI agents.

Why langchain-foxnose?

FoxNose eliminates the "DIY RAG problem" where developers need to stitch together a primary database + full-text search engine + vector database + ETL scripts. With FoxNose, you get:

  • No separate vector DB — embeddings are automatic when fields are marked vectorizable
  • Single _search endpoint — handles text + vector + hybrid + filters in one request
  • No ETL/sync — content changes auto-update embeddings
  • Enterprise features — environments, localization, versioning, RBAC out of the box

This package provides LangChain integrations that wrap the FoxNose Flux API:

  • FoxNoseRetriever — query-based retrieval for RAG pipelines
  • FoxNoseLoader — bulk document loading with cursor-based pagination
  • create_foxnose_tool — search tool for LLM agents

Quick Start

pip install langchain-foxnose

from foxnose_sdk.flux import FluxClient
from foxnose_sdk.auth import SimpleKeyAuth
from langchain_foxnose import FoxNoseRetriever

client = FluxClient(
    base_url="https://<env_key>.fxns.io",
    api_prefix="my_api",
    auth=SimpleKeyAuth("YOUR_PUBLIC_KEY", "YOUR_SECRET_KEY"),
)

retriever = FoxNoseRetriever(
    client=client,
    folder_path="knowledge-base",
    page_content_field="body",
    search_mode="hybrid",
    top_k=5,
)

docs = retriever.invoke("How do I reset my password?")

Next Steps