Web Research Assistant
Learn how to combine the Webpage node with LLM Web Search. Ask a question, and the AI will research using both the reference documentation and real-time web search.
This recipe shows you how to build an app that combines fixed reference pages with LLM web search capabilities. Ask a question, and the AI will research using both the reference documentation and real-time web search to provide comprehensive answers.
Overview
Purpose: Combine fixed reference pages with LLM web search for comprehensive research.
| Node | Role | Model |
|---|---|---|
| App Entry | Question input | - |
| Webpage Node | Fetch reference information from fixed URL | - |
| Generator 1 (Research) | Collect information via web search | claude-haiku-4.5 |
| Generator 2 (Summary) | Summarize research results | gemini-2.5-flash-lite |
What you'll learn: Webpage node, LLM Web Search feature, combining information sources
Workflow Diagram
flowchart LR
A[Research Question] --> C[Web Researcher]
B[Reference Webpage] --> C
C --> D[Summary Generator]Node Configuration Details
1. App Entry
| Item | Value |
|---|---|
| Node Name | Research Question |
| Node Input Name | question |
| Prompt | - (User input only) |
2. Webpage Node
| Item | Value |
|---|---|
| Node Name | Reference Webpage |
| URL | https://docs.giselles.ai/llms-full.txt |
3. Generator Node (Web Researcher)
| Item | Value |
|---|---|
| Node Name | Web Researcher |
| Model | claude-haiku-4.5 |
| Thinking | ON |
| Web Search | ON |
Prompt:
Research the following question using web search and the reference page:
Question: @Research Question
Reference Information: @Reference Webpage
Requirements:
- Search the web for the latest and most relevant information
- Combine web search results with the reference page content
- Provide accurate and up-to-date answers
- Cite sources when possible4. Generator Node (Summary)
| Item | Value |
|---|---|
| Node Name | Summary Generator |
| Model | gemini-2.5-flash-lite |
| Thinking | - |
| Web Search | OFF |
Prompt:
Summarize the research findings into a clear report:
@Web Researcher
Requirements:
- Executive summary (2-3 sentences)
- Key findings as bullet points
- Conclusion and recommendationsBlog Post & Social Media Generator
Learn both sequential and parallel processing in one workflow. Generate blog articles from topics and create optimized social media posts for multiple platforms.
Knowledge Base Q&A
Learn how to build a RAG (Retrieval-Augmented Generation) workflow using both Document Vector Store and GitHub Vector Store together, with query optimization.