mcp-server

smg

Engine-agnostic LLM gateway in Rust. Full OpenAI & Anthropic API compatibility across vLLM, TRT-LLM, TokenSpeed, SGLang, OpenAI, Gemini & more. Industry-first gRPC pipeline, KV cache-aware routing, chat history, tokenization caching, Responses API, embeddings, WASM plugins, MCP, and multi-tenant auth.
572 stars+15★ this week185 forksRustUpdated 10/13/2026100% free · open source
What it does

smg is a Rust‑based gateway that lets you expose any LLM (vLLM, TRT‑LLM, Gemini, etc.) behind a single OpenAI‑compatible API endpoint.

When to use it
  • •You need to swap between different LLM back‑ends without changing your application code.
  • •Your product must support multi‑tenant auth or custom request routing (e.g., per‑customer KV‑cache).
  • •You want to run inference on‑premise but keep the familiar OpenAI HTTP API for your existing stack.
Ready-to-paste prompt
curl http://localhost:8080/v1/chat/completions -H "Content-Type: application/json" -d '{"model":"gpt-4o-mini","messages":[{"role":"system","content":"You are a helpful startup advisor."},{"role":"user","content":"How can I reduce churn for a SaaS product?"}],"temperature":0.7}'
Heads up: smg requires a Rust toolchain ≥ 1.70 and a valid LLM API key (e.g., OPENAI_API_KEY) — the server will start but return 401 for every request if the key is missing or malformed.
Saves to your device
Use with Claude
New

Skip the builder — one click puts this in Claude, Cursor, Antigravity and more.

🔑 Needs an API key

Works after you add credentials — the setup agent will ask you for them.

Try it instantly — no install
Claude Code
mkdir -p ~/.claude/skills/00ea9964-e89e-44c2-85b7-99cc5fc24959 && curl -fsSL https://workflowstacks.com/api/skills/00ea9964-e89e-44c2-85b7-99cc5fc24959/claude-skill -o ~/.claude/skills/00ea9964-e89e-44c2-85b7-99cc5fc24959/SKILL.md
Open in another AI app

Opens the app with this repo with the prompt ready to go — no copy-paste needed.

Connect the whole catalog (MCP)
claude mcp add --transport http workflowstacks https://workflowstacks.com/api/mcp

Adds a WorkflowStacks connector to Claude Code: search and load any skill here by chatting.

How smg works
Codeflow
Free to inspect

Smg is a very large Rust project (~630k lines across 1335 code files, plus 359 test files). Setup is light: installs like a normal app. Reading the code is optional. Last commit this month, Apache-2.0 license, has a test suite.

Size
Very large codebase
~630k lines · 1335 code files · days to read — use, don't read
Setup
One-command install
Installs like a normal app. Reading the code is optional.
Runs on
Rust · Docker
No API keys detected
Rust 85%Python 13%Shell 1%
Where to start reading
  1. 1
    README.md
    Start here — what it does and how to install it
  2. 2
    Cargo.toml
    Dependencies and the commands it exposes
  3. 3
    examples/rl/README.md
    A worked example — copy this to get going
What's in each folder
examples/Examples you can copy23 files
crates/Folder877 files
model_gateway/Folder482 files
grpc_servicer/Folder153 files
e2e_test/Folder123 files
scripts/Helper scripts95 files
bindings/SDKs / client libraries92 files
clients/SDKs / client libraries51 files
READMEHas testsDocumentedCI checksDocker readyExamples includedApache-2.0 licenseUpdated this month
Quick Actions
Details
Creator
smg-project
Language
Rust
Category
mcp-server
Published
11/22/2025

Are you the creator of this tool? and earn 85% of every sale.

Show it off in your README:

Featured on WorkflowStacks
[![Featured on WorkflowStacks](https://workflowstacks.com/api/badge/00ea9964-e89e-44c2-85b7-99cc5fc24959.svg)](https://workflowstacks.com/skills/00ea9964-e89e-44c2-85b7-99cc5fc24959?utm_source=github&utm_medium=badge)
🔥 Hot this week, in your inbox
Skills like this, every Monday.

The five fastest-growing open-source AI skills, ranked by GitHub star growth. One email a week, unsubscribe anytime.