Skip to main content

Web Tools

Prebuilt tools for fetching content from the public web and running Google-powered searches.

Import path: agentflow.prebuilt.tools

fetch_url​

Fetches a public HTTP/HTTPS URL and returns the page content as plain text.

What it does​

  • Resolves the hostname and blocks private/loopback/reserved IP addresses (SSRF protection)
  • Strips HTML tags and script/style content, returning clean readable text
  • Truncates long responses to max_chars (default 20 000)
  • Returns a JSON object with url, status_code, content_type, content, and truncated

Parameters​

ParameterTypeDefaultDescription
urlstrrequiredPublic HTTP or HTTPS URL to fetch
timeoutfloat10.0Request timeout in seconds (clamped 1–30 s)
max_charsint20000Maximum characters to return

Example response​

{
  "url": "https://example.com/",
  "status_code": 200,
  "content_type": "text/html; charset=UTF-8",
  "content": "Example Domain This domain is for use in...",
  "truncated": false
}

Usage​

from agentflow.prebuilt.tools import fetch_url
from agentflow.core.graph import Agent, ToolNode

agent = Agent(
    model="gpt-4o-mini",
    tool_node=ToolNode([fetch_url]),
)

Searches the public web using Gemini Google Search grounding and returns the grounded answer plus source metadata.

What it does​

  • Calls the Google GenAI API with the google_search tool enabled
  • Returns the grounded text answer and grounding_metadata (source links, web chunks)
  • Truncates responses to max_chars

Requirements​

pip install "10xscale-agentflow[google-genai]"

The GOOGLE_API_KEY (or Application Default Credentials) environment variable must be set.

Parameters​

ParameterTypeDefaultDescription
querystrrequiredSearch query
modelstr"gemini-2.5-flash"Gemini model to use
max_charsint20000Maximum characters in the content field

Example response​

{
  "content": "The Eiffel Tower is 330 metres tall...",
  "grounding_metadata": {
    "web_search_queries": ["eiffel tower height"],
    "grounding_chunks": [...]
  },
  "truncated": false
}

Usage​

from agentflow.prebuilt.tools import google_web_search
from agentflow.core.graph import Agent, ToolNode

agent = Agent(
    model="gemini-2.5-flash",
    tool_node=ToolNode([google_web_search]),
)

Searches a Vertex AI Search datastore with Gemini grounding. Suitable for enterprise search over private document collections.

What it does​

  • Calls the Google GenAI API (v1) with a vertex_ai_search retrieval tool
  • The datastore must be a full Vertex AI Search resource path
  • Returns the same content / grounding_metadata / truncated envelope as google_web_search

Requirements​

pip install "10xscale-agentflow[google-genai]"

Vertex AI credentials and a provisioned datastore are required.

Parameters​

ParameterTypeDefaultDescription
querystrrequiredSearch query
datastorestrrequiredFull Vertex AI Search datastore resource path
modelstr"gemini-2.5-flash"Gemini model to use
max_charsint20000Maximum characters in the content field

Usage​

from agentflow.prebuilt.tools import vertex_ai_search
from agentflow.core.graph import Agent, ToolNode

DATASTORE = "projects/my-project/locations/global/collections/default_collection/dataStores/my-store"

agent = Agent(
    model="gemini-2.5-flash",
    tool_node=ToolNode([vertex_ai_search]),
    system_prompt=[{
        "role": "system",
        "content": f"Always search using datastore: {DATASTORE}",
    }],
)

Using multiple web tools together​

from agentflow.prebuilt.tools import fetch_url, google_web_search
from agentflow.core.graph import Agent, ToolNode
from agentflow.prebuilt.agent import ReactAgent

agent = ReactAgent(
    model="gemini-2.5-flash",
    tools=[fetch_url, google_web_search],
    system_prompt=[{
        "role": "system",
        "content": "You are a research assistant. Use google_web_search to find information, "
                   "then fetch_url to read specific pages in full.",
    }],
)
app = agent.compile()
Last updated on