This is the second part of a multipart post about creating a multi-agent research assistant using
The agent had access to a web search tool and would return a list of URL's (and only URL's) that it found related to the topic. However, the long term goal is to add these URL's to a database, and to be able to expand the research later. Because of this, I need to verify that the URL's discovered by the web search agent are not already in a database.
AstraDB
To begin, I created a Datastax AstraDB Serverless database, added a collection and keyspace, and uploaded some sample data. The data consists of 15 URL's that the web search agent had found previously. (For now, I loaded the data manually, but will automate the loop in a later step.)
Sample data
After added the new tool and adjusting the agent's system prompt, the modified flow for agent search with verification looks like this:
Web Search and Verification Agent Flow
At this point, I have a flow that checks my user's input for safety. I also have an agent that returns a list of websites and compares those websites to those already stored in a database. The next step for the multi-agent research assistant will be to summarize the content of each website. That will be Part 3.
SOCIAL SHARE CARD GENERATOR