"Serena" matching MCP servers:

Token-optimized Serena MCP server for AI assistants, reduces context usage by 50% with zero functionality loss.

Enables multiple coding-agent sessions (Claude Code, Codex, and other MCP clients) in the same repository to share a single Serena server and its language servers, instead of each session starting its own. This reduces memory and CPU use, speeds up session startup, and prevents separate instances from overwriting each other's cache.

Bridges stdio MCP clients to shared, warm per-project backend instances over streamable HTTP, so repeated client sessions skip cold starts and transparently survive backend restarts. Heartbeat leases, idle-TTL reaping, and per-request deadlines keep per-project instances alive and shared across multiple clients without hanging the caller.