Before you start
- Access to the Anthale dashboard for the Organization you want to use.
- A Python or Node.js environment, depending on the path you choose.
- One workflow you can test, such as a customer support assistant, a retrieval flow, or an internal copilot.
Recommended integration paths
OpenAI Integration for Python
Wrap the OpenAI Python client with Anthale when your application already uses the OpenAI SDK.
OpenAI Integration for TypeScript
Wrap the OpenAI Node client with Anthale when your application already uses the OpenAI SDK.
LangChain Agent Integration
Add Anthale middleware to a Python LangChain agent workflow with tools and agent execution.
LangChain Chat Integration
Guard a standalone LangChain chat model in Python when you do not need full agent middleware.
Direct SDK paths
Use the direct SDK path when you want explicit control overallow, detect, redact, and block, or when you are building a custom integration Anthale does not already wrap.
Python Direct SDK Path
Send Anthale enforcement requests directly from Python and handle the runtime action in your own application code.
TypeScript Direct SDK Path
Send Anthale enforcement requests directly from TypeScript and handle the runtime action in your own application
code.
Recommended order
- Create an API key for the service that will call Anthale.
- Create a first policy for one trust boundary.
- Start with the integration guide that matches the client or framework already in your application.
- Use the Python Direct SDK Path or TypeScript Direct SDK Path only when you need lower-level control.