ZeroTime/Browser Agent

New — Browser automation for ZeroTime

Give your ZeroTime agents eyes.

Extract structured data from any website, automate web workflows, enrich CRM leads, monitor competitors, and store insights directly in agent memory — all via MCP.

Install browser-mcp

npm · v1.1.1

$ npx @ainative/browser-mcp

Requires AINATIVE_API_KEY — get yours free

0Tools

0Command

0msAvg extract

Free tierEntry plan

See it in action

One command to spin up. One tool call to extract. Watch the agent pull competitor pricing live.

browser-mcp — demo

Up and running in 3 steps

No headless browser infra. No Playwright setup. Just one command.

Install

MCP server starts in seconds

$ npx @ainative/browser-mcp

Connect

Add to your agent config

{
  "mcpServers": {
    "browser": {
      "command": "npx",
      "args": ["@ainative/browser-mcp"]
    }
  }
}

Extract

Your agent browses the web

browser_extract({
  url: "https://target.com",
  extract_goal: "Pricing plans"
})

Built for ZeroTime workflows

Browser Agent plugs directly into every ZeroTime app. Your agents go from passive to proactive.

ZeroPipeline

Auto-enrich CRM leads

Before every sales call, your agent visits the prospect's website, LinkedIn, and Crunchbase — extracts company size, funding, tech stack, and recent news — and stores it directly in the deal record.

browser_enrich_memory({ url: "https://prospect.com", extract_goal: "Company size, funding, tech stack", memory_type: "semantic", project_id: YOUR_PROJECT_ID })

ZeroCommerce

Monitor competitor pricing

Schedule daily extraction of competitor product pages. Results land in a ZeroDB table your agents can query to auto-adjust your own pricing and flag out-of-stock opportunities.

browser_extract_to_table({ url: "https://competitor.com/products", extract_goal: "Product names and prices", table_name: "competitor_pricing", project_id: YOUR_PROJECT_ID })

ZeroInvoice

Extract invoices without APIs

Pull invoice and payment data from client portals, accounting dashboards, or legacy systems that don't expose APIs — extracted into structured JSON, reconciled against your ZeroInvoice records.

browser_extract({ url: "https://clientportal.com/invoices", extract_goal: "Invoice numbers, amounts, due dates" })

OpenCapStack

Research investors before the pitch

Before provisioning a deal room, your agent researches each investor's portfolio, thesis, and recent investments — summarised in the deal room context so you walk in prepared.

browser_enrich_memory({ url: "https://vc-firm.com/portfolio", extract_goal: "Investment thesis and recent portfolio", memory_type: "episodic", project_id: YOUR_PROJECT_ID })

6 tools, 1 install

Every tool is available the moment you run npx @ainative/browser-mcp

Tool	Description	Credits
`browser_act`	Perform actions on any web page — click, type, navigate	50
`browser_extract`	Extract structured data from a page into JSON	75
`browser_validate`	Validate content or state on a page	25
`browser_task`	Run multi-step automation tasks end-to-end	200
`browser_extract_to_table`	Extract data and store directly in a ZeroDB table	100
`browser_enrich_memory`	Extract content from a URL and store in agent memory	100

Browser Agent MCP

Traditional scrapers

Single npx install

Structured JSON output

Works inside MCP agents

Store directly to ZeroDB

No headless infra to manage

Full comparison

Configure in Claude Code

Add to your .claude/mcp.json

.claude/mcp.json

{
  "mcpServers": {
    "browser": {
      "command": "npx",
      "args": ["@ainative/browser-mcp"],
      "env": {
        "AINATIVE_API_KEY": "ak_your_key_here"
      }
    }
  }
}

Works with Claude Code, Cursor, Windsurf, and any MCP-compatible client. Full setup guide →

Simple, transparent pricing

Start free. Scale as your agents grow.

Ready to give your agents eyes?

Free tier. No credit card required. One command to install.

$ npx @ainative/browser-mcp