---
title: "Plug In — Context Infrastructure"
description: "Pick the door that matches where your code already lives."
canonical: "https://press.getalchemystai.com/books/context-infrastructure-build-ai-that-understands-your-business/plug-in"
last_updated: "2018-10-20"
book: "Context Infrastructure"
label: "Chapter 11"
---

# Plug In

Alchemyst is framework-agnostic: it sits beside your orchestration layer rather than inside it. Pick the door that matches where your code already lives.

| Integration | What you get | Notes |
| --- | --- | --- |
| **TypeScript SDK** (`@alchemystai/sdk`) | Typed access to context, memory, async jobs, views and traces | Node 18+, Apache-2.0 |
| **Python SDK** (`alchemystai`) | The same resources, `from alchemyst_ai import AlchemystAI` | Python 3.9+, Apache-2.0 |
| **Vercel AI SDK** | Native middleware for personalised chat, streaming and bulk operations | Uses `sessionId` (formerly `conversationId`) |
| **LangChain, LlamaIndex, n8n** | Context and memory inside existing chains, indexes and workflows | Also works beside LangGraph, CrewAI, Mastra and the OpenAI SDK |
| **MCP server** (`mcp.getalchemystai.com/mcp`) | Context and memory tools inside Claude Desktop, Cursor, VS Code and any MCP client | OAuth 2.1 with PKCE; the API key is entered once at a hosted login, never stored in client config |
| **Context Router** | An OpenAI-compatible proxy that injects scoped context into every inference request | Existing agents gain context with zero code changes |
| **Chrome extension** | Alchemyst context available while browsing | For individual users |
| **Data sources** | Guides for pulling from MongoDB and PostgreSQL | Denormalise relational data before ingesting (Chapter 13) |

**The platform UI** completes the toolkit: a Playground for uploading data and experimenting, a visual context graph with document and node views (nodes can be edited or deleted in place), a trace viewer, and **magic keys** for sharing a context set read-only without re-ingesting it.

**For agents that read docs.** The website exposes machine-readable entry points: `llms.txt`, `llms-full.txt`, an OpenAPI spec at `/openapi.json`, and a separate documentation MCP endpoint at `getalchemystai.com/mcp`. Point a coding agent at these and it can discover the API on its own.

---

- Previous: [Chapter 10: Ask and Steer](https://press.getalchemystai.com/books/context-infrastructure-build-ai-that-understands-your-business/ask-and-steer.md)
- Contents: [Context Infrastructure](https://press.getalchemystai.com/books/context-infrastructure-build-ai-that-understands-your-business.md)
- Next: [Chapter 12: Blueprints](https://press.getalchemystai.com/books/context-infrastructure-build-ai-that-understands-your-business/blueprints.md)
