Search Web for Real-time AI Reasoning
Seltz Web Knowledge Tool delivers context-engineered, real-time web data optimized for LLMs, AI agents, and RAG pipelines through a LlamaIndex integration.
Why it matters
Integrate Seltz's fast, up-to-date web data into your AI reasoning pipelines. Get context-engineered web content and sources to enhance LLMs, AI agents, and RAG systems.
Outcomes
What it gets done
Perform web searches with Seltz.
Retrieve relevant documents with source URLs.
Process web content for AI agent use.
Enhance RAG pipelines with real-time data.
Install
Add it to your toolbox
Run in your project directory:
curl -fsSL https://spark.entire.vc/get/li-tool-tools-seltz | bash Overview
Seltz Web Knowledge Tool
Seltz Web Knowledge Tool searches the web and returns context-engineered content optimized for LLMs, AI agents, and RAG pipelines. It provides a search function that accepts queries and returns Document objects containing web content and source URLs, with optional parameters for maximum documents, context, and profile customization. Use this tool when your AI agent needs current web information for reasoning tasks or when building RAG systems that require fresh, LLM-ready web content. It's ideal for applications where timeliness matters and web data needs to be pre-processed for optimal LLM consumption.
What it does
Seltz Web Knowledge Tool provides fast, up-to-date web data with context-engineered content and sources for real-time AI reasoning. Web content is processed and shaped to maximize usefulness for LLMs, AI agents, and RAG pipelines, returning relevant documents with source URLs through a search interface.
When to use - and when NOT to
Use this tool when your AI agent needs current web information for reasoning tasks, when building RAG pipelines that require fresh web content, or when you need web search results pre-processed for LLM consumption. Use it for queries like "What are the latest developments in AI reasoning?" where timeliness and context matter.
Do NOT use this tool if you need historical web archives or if your application requires offline operation, as it depends on real-time web access and API connectivity.
Inputs and outputs
You provide a search query string, optionally specifying the maximum number of documents to return (default: 10), additional context to refine results, and a profile to customize search behavior. The tool returns a list of Document objects containing web content and source URLs in the metadata.
Integrations
This tool integrates with LlamaIndex as a ToolSpec and works with LlamaIndex FunctionAgent workflows. It is compatible with OpenAI models (demonstrated with gpt-4o).
Who it's for
This tool serves AI developers building agents that need real-time web knowledge, engineers implementing RAG systems with current web content, and teams creating LLM applications where up-to-date information is critical.
Installation and usage
Install via pip:
pip install llama-index-tools-seltz
Basic agent integration:
from llama_index.tools.seltz import SeltzToolSpec
from llama_index.core.agent.workflow import FunctionAgent
from llama_index.llms.openai import OpenAI
seltz_tool = SeltzToolSpec(api_key="your-seltz-api-key")
agent = FunctionAgent(
tools=seltz_tool.to_tool_list(),
llm=OpenAI(model="gpt-4o"),
)
await agent.run("What are the latest developments in AI reasoning?")
Direct search usage:
from llama_index.tools.seltz import SeltzToolSpec
seltz_tool = SeltzToolSpec(api_key="your-seltz-api-key")
documents = seltz_tool.search("web knowledge for AI agents", max_documents=5)
for doc in documents:
print(f"URL: {doc.metadata['url']}")
print(f"Content: {doc.text[:200]}...")
You need to obtain an API key from Seltz before using the tool.
Source README
Seltz Web Knowledge Tool
Seltz provides fast, up-to-date web data with context-engineered web content and sources for real-time AI reasoning. Web content is processed and shaped to maximize usefulness for LLMs, AI agents, and RAG pipelines.
To begin, you need to obtain an API key from Seltz.
Installation
pip install llama-index-tools-seltz
Usage
from llama_index.tools.seltz import SeltzToolSpec
from llama_index.core.agent.workflow import FunctionAgent
from llama_index.llms.openai import OpenAI
seltz_tool = SeltzToolSpec(api_key="your-seltz-api-key")
agent = FunctionAgent(
tools=seltz_tool.to_tool_list(),
llm=OpenAI(model="gpt-4o"),
)
await agent.run("What are the latest developments in AI reasoning?")
Available Functions
search: Search the web using Seltz and return relevant documents with sources. Returns a list of Document objects containing web content and source URLs.
Parameters
query(str): The search query text.max_documents(int, optional): Maximum number of documents to return (default: 10).context(str, optional): Additional context to refine search results.profile(str, optional): Profile to customize search behavior.
Example
from llama_index.tools.seltz import SeltzToolSpec
seltz_tool = SeltzToolSpec(api_key="your-seltz-api-key")
documents = seltz_tool.search("web knowledge for AI agents", max_documents=5)
for doc in documents:
print(f"URL: {doc.metadata['url']}")
print(f"Content: {doc.text[:200]}...")
This tool is designed to be used as a way to load data as a Tool in an Agent.
FAQ
Common questions
Discussion
Questions & comments · 0
Sign In Sign in to leave a comment.