Skip to main content
Use this page when your application already calls the Anthropic JavaScript SDK. The wrapper keeps the Anthropic client shape and adds Averta checkpoints around messages.create(...).

​
Install

npm install @anthropic-ai/sdk @averta-security/sdk-anthropic

export ANTHROPIC_API_KEY="your-anthropic-key"
export AVERTA_API_KEY="your-averta-key"
The Anthropic wrapper reads AVERTA_API_KEY from the environment. Pass avertaApiKey only when this client should use a different Averta key.

​
Wrap the Client

import Anthropic from "@anthropic-ai/sdk";
import { wrapAnthropic } from "@averta-security/sdk-anthropic";

let client = new Anthropic({
  apiKey: process.env.ANTHROPIC_API_KEY!,
});

client = wrapAnthropic(client);

​
Send a Guarded Message

const message = await client.messages.create({
  model: process.env.ANTHROPIC_MODEL ?? "claude-sonnet-4-5",
  max_tokens: 512,
  messages: [
    {
      role: "user",
      content: "Explain password reset link expiration in one sentence.",
    },
  ],
});

console.log(
  message.content
    .map((block) => (block.type === "text" ? block.text : ""))
    .join("\n")
);
For this call, Averta runs a request checkpoint before Anthropic sees the message and an output checkpoint before your app receives the final text.

​
Add Tools

const tools = [
  {
    name: "search_docs",
    description: "Search internal support documentation.",
    input_schema: {
      type: "object",
      properties: {
        query: { type: "string" },
      },
      required: ["query"],
      additionalProperties: false,
    },
  },
];

const toolMessage = await client.messages.create({
  model: process.env.ANTHROPIC_MODEL ?? "claude-sonnet-4-5",
  max_tokens: 512,
  messages: [
    {
      role: "user",
      content: "Search the docs for password reset guidance.",
    },
  ],
  tools,
});
If Anthropic returns tool_use blocks, Averta has already checked those tool calls before your app receives them. When your app sends tool_result blocks back, Averta checks those results before Anthropic sees them.

​
Confirm Decisions

With an allowing policy and a tool call, Events should show a sequence like: 
request -> tool_call -> tool_result -> output
The exact sequence depends on whether the model calls a tool. A request with no tool call can skip tool_call and tool_result. The SDK generates requestId and traceId automatically; pass requestContext only when you need your own conversationId, requestId, or traceId.

​
Runnable Example

The SDK repository includes a complete Anthropic Messages tool-loop example: 
cp examples/.env.example examples/.env
npm --prefix examples/anthropic/basic install
npm run example:anthropic:basic:start
Fill ANTHROPIC_API_KEY and AVERTA_API_KEY in examples/.env before running it.

​
Current Limit

Anthropic streaming is not supported yet. Streaming calls fail closed before provider execution. Use non-streaming messages.create(...) for guarded Anthropic flows today.

​
Next Steps

Messages API

Understand the guarded Anthropic surface.

Tools

Learn how tool exposure and tool-call checks work.

Tool results

Screen returned tool content before continuation calls.

Events

Investigate decisions in the dashboard.