Spike: Decentralized Search and Scrape for Jan-Server #222
Description
Summary
Investigate and prototype low cost search and scraping that does not rely on commercial search APIs. Prioritize decentralized and client driven flows, self hosted or offline indexes, and visible in browser actions that the user consents to. Deliver a PoC plus cost and quality measurements.
Goals
- Return top N candidate pages with title, url, snippet, lightweight score
- Scrape target urls into structured chunks with citations and stable anchors
- Operate without paid search APIs and with minimal central infrastructure
- Integrate with existing MCP tool calls from clients with no schema change