MongoDB

Save Notte browser scrapes to MongoDB in Python or TypeScript.

AI agent instructions

Read the Notte skill and follow its setup instructions when setting up Notte or building browser automation. This is the shared source for CLI setup, authentication, and the browser-to-code workflow.

Save browser content to MongoDB with its official drivers. These examples scrape a page through Notte, insert a document, and close both connections.

Set up

For Python 3.11+, run pip install notte-sdk pymongo==4.19.0. For Node.js 24+, run npm install notte-sdk mongodb@7.7.0 and set "type": "module" in package.json.

Set NOTTE_API_KEY to your Notte API key and MONGODB_URI to your MongoDB connection string. Use a local MongoDB server or an existing Atlas deployment.

Save a page

Save the example as notte_mongodb.py or notte_mongodb.ts. Run python notte_mongodb.py or node --experimental-strip-types notte_mongodb.ts.

notte_mongodb.py

import os
from typing import Any

from notte_sdk import NotteClient
from pymongo import MongoClient
from pymongo.errors import OperationFailure

url = "https://example.com"
uri = os.environ.get("MONGODB_URI")
if not uri:
    raise ValueError("Set MONGODB_URI before running this example")

with NotteClient().Session(max_duration_minutes=3) as session:
    stage = "MongoDB connection"
    try:
        with MongoClient[dict[str, Any]](uri, timeoutMS=10_000, serverSelectionTimeoutMS=10_000) as mongo:
            mongo.admin.command("ping")
            collection = mongo[os.environ.get("MONGODB_DATABASE", "notte")][
                os.environ.get("MONGODB_COLLECTION", "pages")
            ]
            stage = "browser navigation"
            session.execute(type="goto", url=url)
            stage = "browser scrape"
            markdown = session.scrape()
            stage = "MongoDB insert"
            inserted = collection.insert_one({"url": url, "markdown": markdown, "session_id": session.session_id})
            result = {"document_id": str(inserted.inserted_id), "session_id": session.session_id}
            print(result["document_id"])
    except Exception as error:
        code = error.code if isinstance(error, OperationFailure) else None
        raise RuntimeError(
            f"MongoDB export failed (session {session.session_id}, code {code}): {stage}: {type(error).__name__}"
        ) from None

The example prints the inserted document ID. Each document contains the source URL, scraped Markdown, and Notte session ID. Set MONGODB_DATABASE and MONGODB_COLLECTION to choose the destination; they default to notte and pages.

See scraping and session lifecycle.