Byterover

This gives Claude Code access to ByteRover's context engineering platform, which is basically a structured way to store and query project knowledge instead of scattered markdown files. You get organized domains like database schema, testing strategies, and deployment patterns that sync across your team automatically. The skill runs through Docker since ByteRover needs Node.js 20+ and credential storage, and it supports headless mode with JSON output so agents can actually use it. The main workflow is curate to add context, query to retrieve it, and push/pull to sync. Honest take: if you're tired of maintaining .cursorrules or context.md files manually, this automates that whole dance, though it does require signing up for their service.