---

  You give it a task
        │
        ▼
  Opens Bright Data Scraping Browser (real Chromium, residential IPs)
        │
        ▼
  Takes a screenshot → sends to Yutori n1 model
        │
        ▼
  n1 reasons: click here, type this, scroll there...
        │
        ▼
  Repeats until confident enough to answer
        │
        ▼
  Returns result to your terminal

---

Before you begin, you'll need accounts and credentials from two services:

Service	What you need	Where to get it
Bright Data	Scraping Browser CDP URL	brightdata.com → Scraping Browser zone
Yutori	API Key	yutori.com

---

1. Clone the repository

git clone https://github.com/meirk-brd/n1-brightdata.git
cd n1-brightdata

2. (Recommended) Create a virtual environment

python -m venv .venv
source .venv/bin/activate       # macOS / Linux
# .venv\Scripts\activate        # Windows

3. Install the CLI

pip install -e .

This installs the n1-brightdata command globally in your environment. Because it's editable (-e), any local code changes take effect immediately — no reinstall needed.

4. Install the browser

playwright install chromium

---

Option A — Interactive Setup (Recommended)

Run the built-in setup wizard. It walks you through every credential, saves them to .env, installs Playwright, and optionally tests your connections:

n1-brightdata setup

Option B — Manual credentials.json

Create the file at ~/.n1-brightdata/credentials.json:

mkdir -p ~/.n1-brightdata
cat > ~/.n1-brightdata/credentials.json << 'EOF'
{
  "YUTORI_API_KEY": "your_yutori_api_key_here",
  "BRD_CDP_URL": "wss://brd-customer-XXXXXX-zone-scraping_browser:PASSWORD@brd.superproxy.io:9222"
}
EOF
chmod 600 ~/.n1-brightdata/credentials.json

Optional Tuning (project-level)

Tuning parameters stay in a local .env file or shell environment — they are not stored in credentials.json:

# .env  (in your project directory, or export in your shell)
N1_SCREENSHOT_FORMAT=jpeg          # jpeg | png
N1_JPEG_QUALITY=60                 # 1–100
N1_SCREENSHOT_TIMEOUT_MS=90000     # screenshot timeout in milliseconds
N1_MAX_REQUEST_BYTES=9500000       # trim old screenshots if payload exceeds this
N1_KEEP_RECENT_SCREENSHOTS=6       # how many screenshots to keep in context
N1_ENABLE_SUFFICIENCY_CHECK=true   # stop early when the agent is confident
N1_STOP_CONFIDENCE_THRESHOLD=0.78  # confidence threshold for early stopping (0–1)

Configuration Precedence

CLI flags  >  Shell env vars  >  .env file (tuning)  >  ~/.n1-brightdata/credentials.json

---

Basic

n1-brightdata "What is the current price of Bitcoin?"

With a Starting URL

n1-brightdata "Find the top 5 trending repositories" --url "https://github.com/trending"

Limit the Number of Steps

n1-brightdata "Summarize today's top news" --url "https://news.ycombinator.com" --max-steps 15

Full Command Reference

n1-brightdata [TASK] [OPTIONS]

Option	Type	Default	Description
TASK	string	(required)	The task for the agent to complete
--url	string	https://www.google.com	Starting URL before the agent loop begins
--max-steps	integer	30	Maximum number of browser actions (min: 1)
--screenshot-format	jpeg | png	jpeg	Format of screenshots sent to the model
--jpeg-quality	integer	60	JPEG quality when format is jpeg (1–100)
--screenshot-timeout-ms	integer	90000	Screenshot timeout in milliseconds
--show-inspect-url	flag	false	Print a DevTools inspect URL for the live session
--yutori-api-key	string	(env)	Yutori API key (overrides env / .env)
--brd-cdp-url	string	(env)	Bright Data CDP WebSocket URL (overrides env / .env)
--env-file	path	./.env	Custom path to a .env credentials file

Help

n1-brightdata --help
n1-brightdata run --help
n1-brightdata setup --help

---

# Research a topic
n1-brightdata "What are the key differences between GPT-4o and Claude 3.5 Sonnet?"

# Scrape structured data
n1-brightdata "List the top 10 products on Product Hunt today" --url "https://www.producthunt.com"

# Monitor a specific page
n1-brightdata "Is there a sale on the MacBook Pro 16-inch?" \
  --url "https://www.apple.com/shop/buy-mac/macbook-pro" \
  --max-steps 10

# High-quality screenshots for visual tasks
n1-brightdata "Describe the layout of the Airbnb homepage" \
  --url "https://www.airbnb.com" \
  --screenshot-format png

# Use a different .env for multiple accounts
n1-brightdata "Check order status" --env-file ~/.config/n1/work.env

---

n1-brightdata/
├── src/
│   └── n1_brightdata/
│       ├── __init__.py      # Package exports: AgentConfig, build_agent_config, run_agent
│       ├── agent.py         # Core agentic loop, browser tools, n1 model integration
│       ├── cli.py           # Click CLI: `run` and `setup` commands
│       └── console.py       # Rich terminal UI: banners, step display, progress
├── pyproject.toml           # Project metadata, dependencies, CLI entrypoint
└── .env                     # Your credentials (not committed)

---

---

---