@awesome-agent/adapter-anthropic

Anthropic Claude adapter for @awesome-agent/agent-core.

Uses the native Anthropic Messages API — not the OpenAI compatibility layer. Supports streaming, tool use, and all Claude models.

Installation

npm install @awesome-agent/agent-core @awesome-agent/adapter-anthropic

Usage

import { AnthropicAdapter } from "@awesome-agent/adapter-anthropic";

const llm = new AnthropicAdapter({
  apiKey: process.env.ANTHROPIC_API_KEY!,
});

With Custom Base URL (proxy)

const llm = new AnthropicAdapter({
  apiKey: process.env.ANTHROPIC_API_KEY!,
  baseURL: "https://my-proxy.com",
});

With Retry

import { RetryLLMAdapter } from "@awesome-agent/agent-core";
import { AnthropicAdapter } from "@awesome-agent/adapter-anthropic";

const llm = new RetryLLMAdapter(
  new AnthropicAdapter({ apiKey: process.env.ANTHROPIC_API_KEY! }),
  { maxRetries: 3 }
);

Full Example with AgenticLoop

import {
  AgenticLoop, DefaultToolRegistry, DefaultToolExecutor,
  DefaultHookManager, DefaultContextBuilder,
} from "@awesome-agent/agent-core";
import { AnthropicAdapter } from "@awesome-agent/adapter-anthropic";

const llm = new AnthropicAdapter({
  apiKey: process.env.ANTHROPIC_API_KEY!,
});

const tools = new DefaultToolRegistry();
const loop = new AgenticLoop({
  llm,
  agent: {
    id: "claude-agent",
    name: "Claude Agent",
    prompt: "You are a helpful assistant.",
    model: "claude-sonnet-4-20250514",
  },
  tools,
  executor: new DefaultToolExecutor(tools),
  hooks: new DefaultHookManager(),
  context: new DefaultContextBuilder(),
});

const result = await loop.run("Hello Claude!", "session-1");
console.log(result.output);

API Differences from OpenAI

This adapter handles Anthropic's unique API format automatically:

| Feature | OpenAI | Anthropic (handled by adapter) | |---------|--------|-------------------------------| | System prompt | Message with role: "system" | Top-level system parameter | | Tool calls | function type | tool_use type | | Tool results | role: "tool" message | tool_result content block in user message | | Streaming events | choices[0].delta | content_block_delta, message_delta | | Stop reason | finish_reason: "stop" | stop_reason: "end_turn" |

You don't need to handle any of this — just use the adapter like any other LLMAdapter.

Configuration

| Option | Type | Default | Description | |--------|------|---------|-------------| | apiKey | string | (required) | Anthropic API key | | baseURL | string | "https://api.anthropic.com" | API base URL | | apiVersion | string | "2023-06-01" | anthropic-version header | | defaultHeaders | Record<string, string> | {} | Extra headers |

Exports

| Export | Description | |--------|-------------| | AnthropicAdapter | LLMAdapter implementation for Anthropic Messages API | | AnthropicAdapterConfig | Configuration type | | AnthropicStreamParser | SSE parser for Anthropic stream format (advanced) |

License

Apache-2.0