Autonomous Research & Library Agent

I need an AI-driven agent that actively grows my digital library by roaming the open web, identifying valuable material from discussion forums, blogs, Q&A sites and similar sources, extracting that content and its metadata, then storing everything in a structured library collection. The goal is not traditional cataloging or loan management; the agent’s core job is continuous research and ingestion. The end results should look more like profile that is easy to digest.

Here’s how I picture the workflow. The agent receives a topic list or keyword feed, searches the public internet, screens the results for relevance, collects the full text along with author, date, source URL and any tags it can infer, removes duplicates, normalises the data and finally inserts the new records into my library database or a well-documented local store that I can use for content purposes and later on ingegrate with other systems (like chatbot). Robust error handling, rate-limit compliance and source attribution are essential. Verification of data by admin is quite must have. This is not fixed plan and open for improvements.

I already have SmythOS subscription, which would be the most optimal if it can be useful. Otherwise explain why to build whose system. First version needs just to give an insight that this agent can really to good job in that category and required accuracy.

If this sounds within your wheelhouse, send a note outlining your proposed architecture and any similar autonomous agents you’ve built. Looking for most efficient solution at the beginning to get the sense how it works. Please tell some estimates for work load for such a prototype.