Auto-Discovery overview

Auto-Discovery as a cross-cutting topic — automated discovery of new tools, MCP servers, frameworks, and libraries from npm registry trending data and GitHub trending repositories for graph enrichment. Covers polling strategies (npm trending API, GitHub trending scraping, release feed monitoring), signal filtering (filter noise from star counts vs genuine adoption signals), deduplication against already-known nodes, candidate scoring (stars, weekly downloads, MCP-compatibility), and the human-in-the-loop review step before auto-generated nodes are committed to the graph. Auto-discovery closes the gap between the ecosystem's rate of change and the graph's coverage without requiring full manual curation for every new tool release.