Lightweight Plugin Library

Alchymos SDK — add caching & observability in minutes

Thin, dependency‑free SDKs that wrap LLM and MCP clients to deliver edge caching, metrics, and policy controls.

npm install @alchymos/openai

SDKs for OpenAI, MCP clients and servers that provide automatic caching, request/response instrumentation, scoped privacy, and programmatic purge APIs with TypeScript-first ergonomics.

OpenAI

Drop-in wrapper that enables cache + telemetry instantly. Works with your existing OpenAI client configuration.

import OpenAI from "openai";
import { Alchymos } from '@alchymos/openai';

const openai = new OpenAI(process.env.OPENAI_API_KEY);
const alchymos = new Alchymos(process.env.ALCHYMOS_API_KEY);

alchymos.withLlm(openai);

MCP Client

Instrument client-side tool calls for caching and metrics. Seamlessly integrates with the Model Context Protocol.

import { Client } from "@modelcontextprotocol/sdk/client/index.js";
import { Alchymos } from '@alchymos/mcp';

const client = new Client();
const alchymos = new Alchymos(process.env.ALCHYMOS_API_KEY);

alchymos.withClient(client);

MCP Server

Server middleware for edge caching and observability. Add caching capabilities to your tools with one line.

import { Server } from "@modelcontextprotocol/sdk/server/index.js";
import { Alchymos } from '@alchymos/mcp';

const server = new Server();
const alchymos = new Alchymos(process.env.ALCHYMOS_API_KEY);

alchymos.withServer(server);

Full control in TypeScript

Our SDK is designed for modern TypeScript environments. It includes full type definitions, zero runtime dependencies, and flexible configuration options for granular control.

Zero runtime deps

Keep your bundle size small and fast.

TypeScript first

Full type safety and autocompletion.

Scoped Privacy

Cache scoped to specific orgs or users.

Local Dev Mode

Replay logs locally without hitting API.

example.ts
import OpenAI from "openai";
import { Alchymos } from '@alchymos/openai';

const openai = new OpenAI(process.env.OPENAI_API_KEY);
const alchymos = new Alchymos(process.env.ALCHYMOS_API_KEY);

alchymos.withLlm(openai);

const response = await openai.chat.completions.create(params);