ClawQL has friends — MCP + gRPC ecosystem

ClawQL has friends! See how it fits into the wider MCP ecosystem and how we're helping everyone with our new mcp-grpc-transport 👀

MCP + gRPC momentum

Google Cloud's January 2026 announcement of pluggable gRPC transports for MCP marked a turning point for enterprises already standardized on gRPC microservices. The community responded rapidly with high-quality, language-specific tools that make MCP feel truly cloud-native.

TypeScript / Node.js

  • mcp-grpc-transport (by the ClawQL team) — Our pluggable gRPC transport for any @modelcontextprotocol/sdk McpServer. Features: unary + bidirectional streaming, built-in health checks (grpc.health.v1.Health), optional server reflection, mTLS via env vars, and one-line integration. Perfect for long-running agent loops (memory recall, sandbox execution, multi-turn tool composition) and native mesh routing (Cloudflare Mesh, Istio, etc.). → npm | GitHub
  • ClawQL Unified Server — Single-process MCP server with embedded in-process GraphQL proxy for 99%+ token savings in execute() calls. Fully compatible with mcp-grpc-transport for side-by-side stdio/HTTP/gRPC operation.

Go

  • protoc-gen-go-mcp (Redpanda Data) — A Protocol Buffers compiler plugin that auto-generates full MCP servers from any existing gRPC or ConnectRPC service. Automatically creates JSON Schema for tool inputs from your .proto definitions and wires handlers directly to your gRPC backend.
  • gRPC-buf (by Dipjyoti Metia) — A production-ready Go reference implementation featuring dual gRPC + REST APIs (via Connect), PostgreSQL persistence, and comprehensive operational documentation (architecture, config, workflows, playbooks). Excellent blueprint for running MCP + gRPC services at scale.

Why this matters for ClawQL users

Whether your backend team works in Go (exposing existing services via protoc-gen-go-mcp) or TypeScript (using ClawQL + mcp-grpc-transport), you get a consistent, high-performance MCP layer. ClawQL shines as the sovereign TypeScript hub:

  • Keeps core inference and memory fully on-prem in your private AI lab
  • Delivers massive token efficiency through its in-process GraphQL proxy
  • Provides human-inspectable Obsidian knowledge graphs with Merkle-proven provenance (Pro/Enterprise)
  • Supports selective hybrid Cloudflare Mesh/Artifacts usage only for high-value paths

This multi-language interoperability reduces protocol friction and lets teams reuse existing gRPC infrastructure while gaining powerful agentic capabilities. We actively collaborate with and celebrate the broader ecosystem. If you're building complementary tools (e.g., MCP clients, domain-specific connectors, or observability integrations), we'd love to feature them here or explore co-marketing.

Want to integrate? Start with mcp-grpc-transport today — it's completely generic and works with any MCP server, including ClawQL. Questions or ideas for deeper integrations? Open a discussion on GitHub or reach out via the community channels.

Shout out to Redpanda Data and Google and Dipjyoti Metia for their contributions to gRPC + MCP ecosystem. Sincerely hope my typescript based grpc transport is a welcome addition and you all benefit from it as I have from your work. Thank you!

Original thread (comments, reactions, and any media): View on LinkedIn