带视觉能力的智能体

创建能够处理和分析图像的智能体。

快速开始#

克隆完整示例#

最简单的开始方式是克隆完整的示例仓库:

bash
git clone https://github.com/astreus-ai/examples
cd examples/agent-with-vision
npm install

或仅安装包#

如果你更喜欢从零开始构建:

bash
npm install @astreus-ai/astreus

环境配置#

bash
# .env
# Vision-capable model API key
OPENAI_API_KEY=sk-your-openai-api-key-here

# Database for agent persistence
DB_URL=sqlite://./astreus.db

视觉智能体#

typescript
import { Agent } from '@astreus-ai/astreus';

const agent = await Agent.create({
  name: 'VisionBot',
  model: 'gpt-6-astra',
  visionModel: 'gpt-6-astra',
  vision: true,
  systemPrompt: 'You can analyze and describe images in detail.'
});

// Analyze an image
const result = await agent.ask("Analyze this image and describe what you see", {
  attachments: [{
    type: 'image',
    path: './screenshot.png'
  }]
});

console.log(result); // Detailed image analysis

运行示例#

如果你克隆了仓库:

bash
npm run dev

仓库#

完整示例可在 GitHub 上获取:examples/agent-with-vision