Outline is an open-source collaborative knowledge base platform designed for team information sharing.
This notebook shows how to obtain langchain Documents from your Outline collections.

​
Overview

The Outline Document Loader can be used to load Outline collections as LangChain Documents for integration into Retrieval-Augmented Generation (RAG) workflows. This example demonstrates:
  • Setting up a Document Loader to load all documents from an Outline instance.

​
Setup

Before starting, ensure you have the following environment variables set: 
import os

os.environ["OUTLINE_API_KEY"] = "ol_api_xyz123"
os.environ["OUTLINE_INSTANCE_URL"] = "https://app.getoutline.com"

​
Initialization

To initialize the OutlineLoader, you need the following parameters:
  • outline_base_url: The URL of your outline instance (or it will be taken from the environment variable).
  • outline_api_key: Your API key for authenticating with your Outline instance (or it will be taken from the environment variable).
  • outline_collection_id_list: List of collection ids to be retrieved. If None all will be retrieved.
  • page_size: Because the Outline API uses paginated results you can configure how many results (documents) per page will be retrieved per API request. If this is not specified a default will be used.

​
Instantiation

# Option 1: Using environment variables (ensure they are set)
from langchain_outline.document_loaders.outline import OutlineLoader

loader = OutlineLoader()

# Option 2: Passing parameters directly
loader = OutlineLoader(
    outline_base_url="YOUR_OUTLINE_URL", outline_api_key="YOUR_API_KEY"
)

​
Load

To load and return all documents available in the Outline instance 
loader.load()

​
Lazy Load

The lazy_load method allows you to iteratively load documents from the Outline collection, yielding each document as it is fetched: 
loader.lazy_load()

​
API reference

For detailed documentation of all Outline features and configurations head to the API reference: https://www.getoutline.com/developers