Gemini agents that read the web
Agents that fetch and extract before they reason.
Extract the specs table from these three product pages.
MCPToolset with StreamableHTTPConnectionParams gives an ADK agent the full Import.io Web Scraper MCP toolset in a few lines.
import os
from google.adk.agents import LlmAgent
from google.adk.tools.mcp_tool.mcp_toolset import MCPToolset
from google.adk.tools.mcp_tool.mcp_session_manager import StreamableHTTPConnectionParams
root_agent = LlmAgent(
model="gemini-flash-latest",
name="web_data_agent",
instruction="Use Import.io tools to read live pages and return structured results.",
tools=[MCPToolset(connection_params=StreamableHTTPConnectionParams(
url="https://mcp.import.io/mcp",
headers={"Authorization": f"Bearer {os.environ['IMPORTIO_MCP_KEY']}"},
))],
)Any public page, rendered, through managed proxies
Rendering, proxy routing, captcha handling, extraction
https://mcp.import.io/mcp
Import.io as an MCPToolset for Agent Development Kit agents.
Agents that fetch and extract before they reason.
Extract the specs table from these three product pages.
A dedicated web data sub-agent that others delegate to.
Find current prices for this list of SKUs.
The same config works locally and in your deployed agent.
Monitor this page daily and flag changes.
pip install google-adk.https://mcp.import.io/mcp and a bearer header.adk run or adk web.Yes. MCPToolset accepts a tool_filter list; pass the Import.io tool names the agent should see.
Signup includes 10,000 successful calls. After that it is $0.0002 per successful call. Setup requests, tool listings, malformed requests, blocked requests and failed upstream calls are never billed, and you can set a monthly spend limit that stops paid usage when reached.
Use OAuth in interactive clients that support it: the client opens a sign-in, stores the credentials and reuses them. Use an API key for scripts, servers and clients without OAuth, sent as Authorization: Bearer mcp_live_…. Keep keys out of prompts and shared config files.
importio_web_fetch, importio_render, importio_extract_data and importio_get_html, page helpers such as importio_page_text, importio_page_links and importio_search_cards, and browser actions for clicks, form input, scrolling, pagination, waits, screenshots and captcha handling. Requests can pick a proxy country and type, with bounded retries when a page blocks.
Load Import.io tools with langchain-mcp-adapters into any agent.
Setup 10 minMCPGive every agent in a crew live web data tools.
Setup 10 minMCPCall Import.io tools directly from the official Python and TypeScript SDKs.
Setup 10 minMCPLive page data and extraction tools in Google’s terminal agent.
Setup 2 min