## Summary
umerkay/mike
umerkay's fork wires Mike up to live web research, giving the assistant a way to fetch, search, and crawl pages on demand.
This is a fork of Mike maintained by umerkay, and so far the work has a single clear thread: giving Mike a backend pathway to the open web. Where stock Mike reasons over what it already has, this fork starts to ask what happens when the assistant can also go and look something up.
The direction is research-flavoured rather than UI-flavoured. There's nothing here yet about a new chat experience, a rebrand, or a vertical niche - the fork is building plumbing, and the plumbing points outward at third-party web data.
It's early. One thread of work, one commit ahead of upstream, and no public signal yet about who the intended user is or how the new capability surfaces in the product. Worth a click-through if you're interested in how a legal-tech assistant might be extended with live retrieval; less interesting if you're shopping for a finished vertical product.
What's in it
- Web scraping on demand A path for Mike's backend to pull a URL and get back clean, readable content rather than raw HTML.
- Question-answering against a page Point the assistant at a specific URL and ask a question about what's on it, instead of having to ingest the whole thing first.
- Multi-page crawling Kick off a crawl across a site as a background job and collect the results once it's done.
- Site mapping Discover the shape of a site - what URLs exist - before deciding what to actually fetch.
Direction
integrationsearch
Activity
Four new endpoints let the backend scrape pages, ask questions of a URL, crawl whole sites, and map their structure.
Threads of work (detailed view)
umerkay bolts a web research desk onto Mike
Four new endpoints let the backend scrape pages, ask questions of a URL, crawl whole sites, and map their structure.
Pull requests (detailed view)
⛔ Closed without merge (1)
by
umerkay · opened 25d ago · closed 16d ago