Browser Automation

Two browser modes — cloud and local — so agents can operate the web exactly like a person

What is Browser Automation?

Opulent provides two complementary browser environments for autonomous web operation:

Cloud Browser Infrastructure — A fully managed, highly available browser engine in the cloud. The agent browses, clicks, fills forms, extracts data, and executes multi-step workflows — even on complex, JavaScript-heavy single-page applications.
Local Browser Operator — The agent operates your own browser directly, using your active sessions and existing logins. No CAPTCHAs, no re-authentication.

Together, they make Opulent a true web agent — not a chatbot that describes the web, but a system that acts on it.

Cloud Browser Infrastructure provides a dedicated, isolated browser environment. The agent navigates websites, interacts with elements, and completes workflows — all in a managed container. It handles dynamic content seamlessly and does not use your local machine.

Why it Matters

Most AI tools can only work with information you paste in. They cannot visit websites, interact with web applications, or access content behind logins.

Cloud Browser enables Opulent to:

Complete multi-step workflows across many web pages
Extract data from websites that have no API
Interact with web applications as a human would
Browse the web autonomously to gather real-time information
Log into accounts to perform authenticated actions

How It Works

1. Agent Opens the Browser

When you assign a task requiring web access, Opulent automatically provisions a Cloud Browser instance and navigates to the relevant sites.

2. Authenticated Sessions

Log into accounts inside Cloud Browser. The agent uses those authenticated sessions to perform actions on your behalf — updating your CRM, reading emails, posting content, pulling reports.

3. Take Over When Needed

For SMS codes, CAPTCHAs, or MFA challenges, you receive a prompt to take over temporarily. Complete the verification, then hand control back to the agent.

Security and Privacy

Isolated per-user browser instance
Encrypted sessions end-to-end
No password storage
You control which accounts the agent can access
Full audit log of every navigation and action

Accounts to Log Into for Maximum Power

Productivity & Communication

Gmail / Outlook — Read, send, organize emails
Google Calendar — Schedule meetings, check availability
Notion — Create pages, update databases
Slack — Post messages, read channels

Business & CRM

Salesforce / HubSpot — Update leads, log activities, create records
Stripe — Check payments, generate invoices, review subscriptions
LinkedIn — Send requests, read messages, monitor mentions

Research & Data

Crunchbase / PitchBook — Company and funding data
SimilarWeb — Traffic and competitive analysis
Financial Times / WSJ — Premium news

Development

GitHub — Create issues, review PRs, manage repos
Jira / Linear — Update tickets, track sprints

Local Browser Operator

What it is

The Local Browser Operator connects to your desktop browser directly. The agent operates your active browser session — using your existing logins, cookies, and local state — without needing to re-authenticate on any site.

Why Use Local Instead of Cloud

Cloud Browser uses data center IP addresses. Some sites detect this and trigger extra verification, block access, or require re-login. The Local Browser Operator uses your actual IP address and your existing sessions, which means:

No re-authentication needed
No CAPTCHA challenges on sites you're already logged into
Works on banking, financial, and security-sensitive tools
Seamless access to internal tools that require corporate network/VPN

How It Works

Enable: Go to Settings → Browser → Local Browser Operator → Install the browser extension
Grant Access: When you assign a task requiring local browser access, approve the session
Autonomous Action: The agent works in a dedicated tab using your logged-in sessions

Choosing Between Cloud and Local

Use Cloud Browser Infrastructure When	Use Local Browser Operator When
Researching public websites	Logging into security-sensitive accounts
Extracting data at scale	Banking and financial sites
Running long-horizon tasks in background	Sites that block data center IPs
Research doesn't require your logins	Avoiding CAPTCHA/MFA challenges
You want a highly parallelizable environment	You need your residential IP

Key Capabilities (Both Modes)

Navigate and Extract

"Go to these 30 SaaS pricing pages and extract: plan names, prices,
features per tier, and free trial availability. Return as a CSV."

Interact with Web Applications

"Log into our HubSpot account, find all deals in 'Negotiation' stage
that haven't been updated in 14 days, and add a follow-up task to each"

Multi-Step Workflows

"Go to LinkedIn, search for VP of Sales at Series B SaaS companies
in North America, collect the first 50 profiles, and extract:
name, company, title, connection degree, and a direct message link"

Authenticated Actions

"Log into our Stripe account and generate a monthly revenue report
for Q4 — total MRR, new MRR, churned MRR, and top 10 customers by spend"

Form Filling and Submission

"Fill out and submit the vendor onboarding form at this URL using
the information in this spreadsheet — one submission per row"

Common Use Cases

Use Case	Browser Mode	What the Agent Does
Competitive pricing research	Cloud	Visits 50 competitor pages, extracts pricing
CRM data entry	Local	Logs into Salesforce, creates/updates records
LinkedIn outreach	Local	Sends connection requests, reads inbox
Market research	Cloud	Scrapes 100+ pages simultaneously
Banking statement download	Local	Uses your local session, no re-auth
News monitoring	Cloud	Visits news sites, extracts and summarizes
Job board monitoring	Cloud	Tracks new postings across 10 boards
Internal admin tools	Local	Operates tools behind corporate VPN

Tips

Be specific about the desired output format:

✅ "Return as a CSV with columns: company, pricing tier, price, feature list"
❌ "Get me the pricing info"

Use Local Browser for sensitive accounts:

✅ "Log into my banking portal and download last month's statements" → Local Browser
✅ "Research 50 public company websites and extract pricing" → Cloud Browser

Monitor complex workflows: Keep the Browser panel open for multi-step tasks so you can take over immediately if a verification appears.

Log in ahead of time: For Cloud Browser authenticated tasks, log into your accounts once. Sessions persist so the agent can access them without interrupting future workflows.

Common Questions

Is it secure to let the agent use my accounts? Yes. Sessions are encrypted and isolated. The agent cannot access accounts you haven't explicitly logged into. You can revoke access at any time from Settings → Browser.

What happens if the agent hits a CAPTCHA? You'll receive a "Take Over" prompt. Complete the CAPTCHA, then hand control back to the agent.

Can I see what the agent is doing in real time? Yes. The Browser workbench panel shows the live browser session as it operates.

Can the agent handle JavaScript-heavy single-page apps? Yes. Opulent uses highly capable cloud browser infrastructure that natively executes JavaScript and interacts with the DOM exactly as a human would, rather than relying on simple HTTP scraping. Look, read, and act loops ensure full comprehension of dynamic pages.

Can I stop a browser task mid-execution? Yes. Hit Stop in the chat input at any time. The task halts immediately.