CozeLoop Java SDK

English | 简体中文

Java SDK for interacting with CozeLoop platform.

Features

• Trace Reporting: Built on OpenTelemetry SDK, automatic batch reporting
  • Two-level batching: OpenTelemetry BatchSpanProcessor + custom 25-span batches
  • Automatic context propagation across threads and async operations
  • Support for Events, Links, and Baggage
  • See OpenTelemetry Integration Guide for details
• Prompt Management: Pull, cache, format and execute prompts
• AOP Annotation: Declarative tracing with @CozeTrace annotation
• Spring Boot Integration: Seamless integration with Spring Boot applications

Requirements

• Java 8+ (Java 11+ recommended)
• Maven 3.6+

Installation

Maven

<dependency>
    <groupId>com.coze</groupId>
    <artifactId>cozeloop-core</artifactId>
    <version>0.1.0</version>
</dependency>

Spring Boot Starter

<dependency>
    <groupId>com.coze</groupId>
    <artifactId>cozeloop-spring-boot-starter</artifactId>
    <version>0.1.0</version>
</dependency>

Quick Start

Initialize

First, visit https://loop.coze.cn/console/enterprise/personal/open/oauth/apps and create an OAuth app. Then you can get your owner appid, public key and private key.

Set your environment variables:

export COZELOOP_WORKSPACE_ID=your workspace id
export COZELOOP_JWT_OAUTH_CLIENT_ID=your client id
export COZELOOP_JWT_OAUTH_PRIVATE_KEY=your private key
export COZELOOP_JWT_OAUTH_PUBLIC_KEY_ID=your public key id

Or use PAT (Personal Access Token) for testing:

export COZELOOP_WORKSPACE_ID=your workspace id
export COZELOOP_API_TOKEN=your token

Basic Usage

import com.coze.loop.client.CozeLoopClient;
import com.coze.loop.client.CozeLoopClientBuilder;
import com.coze.loop.trace.CozeLoopSpan;
import com.coze.loop.entity.Prompt;
import com.coze.loop.entity.Message;
import com.coze.loop.prompt.GetPromptParam;
import java.util.Map;
import java.util.HashMap;
import java.util.List;

// Initialize client
String workspaceId = System.getenv("COZELOOP_WORKSPACE_ID");
String apiToken = System.getenv("COZELOOP_API_TOKEN");

CozeLoopClient client = new CozeLoopClientBuilder()
    .workspaceId(workspaceId)
    .tokenAuth(apiToken)  // or use .jwtOAuth(clientId, privateKey, publicKeyId)
    .build();

try {
    // Report trace
    try (CozeLoopSpan span = client.startSpan("operation", "custom")) {
        span.setInput("Hello");
        span.setOutput("World");
        span.setAttribute("custom_key", "custom_value");
    }
    
    // Get and format prompt
    Prompt prompt = client.getPrompt(GetPromptParam.builder()
        .promptKey("your_prompt_key")
        .build());
    
    Map<String, Object> variables = new HashMap<>();
    variables.put("var1", "content");
    List<Message> messages = client.formatPrompt(prompt, variables);
} finally {
    // Close client (important: ensures all traces are reported)
    client.close();
}

Report Trace with LLM

try (CozeLoopSpan span = client.startSpan("llmCall", "llm")) {
    String input = "What is the weather in Shanghai?";
    
    // Call your LLM API here
    String output = callLLMAPI(input);
    
    // Set span attributes
    span.setInput(input);
    span.setOutput(output);
    span.setModelProvider("openai");
    span.setModel("gpt-4o-2024-05-13");
    span.setInputTokens(11);
    span.setOutputTokens(52);
}

Spring Boot Integration

application.yml

cozeloop:
  workspace-id: your_workspace_id
  jwt:
    client-id: your_client_id
    private-key: your_private_key
    public-key-id: your_public_key_id
  trace:
    enabled: true

Use Annotations

import com.coze.loop.spring.annotation.CozeTrace;
import org.springframework.stereotype.Service;

@Service
public class LLMService {
    
    @CozeTrace(value = "llm_call", spanType = "llm", captureArgs = true, captureReturn = true)
    public String callLLM(String prompt) {
        return llmClient.call(prompt);
    }
}

Documentation

• Examples - Code examples demonstrating SDK usage
• OpenTelemetry Integration Guide - Detailed guide on OpenTelemetry integration
  • Architecture overview
  • Two-level batching strategy
  • Context propagation
  • Advanced features (Events, Links, Baggage)
  • Best practices
• API Documentation - API reference
• Configuration - Configuration options

OpenTelemetry Benefits

The SDK is built on OpenTelemetry, providing:

• Industry Standard: Widely adopted observability framework
• Vendor Neutral: Works with any backend that supports OpenTelemetry
• Rich Ecosystem: Extensive instrumentation libraries
• Automatic Batching: Built-in batch processing for efficient export
• Context Propagation: Automatic trace context propagation across services
• Mature & Battle-Tested: Production-ready with excellent performance

For more details, see the OpenTelemetry Integration Guide.

Building from Source

git clone https://github.com/coze-dev/cozeloop-java.git
cd cozeloop-java
mvn clean install

Contribution

Please check Contributing for more details.

Security

If you discover a potential security issue in this project, or think you may have discovered a security issue, we ask that you notify Bytedance Security via our security center or vulnerability reporting email.

Please do not create a public GitHub issue.

License

This project is licensed under the MIT License.