AI Agent Workflows in Python: Power Your guild-packs Agents with Multimodal Inference via NexaAPI

# Install: pip install nexaapi guild-packs
from nexaapi import NexaAPI
import os

client = NexaAPI(api_key=os.environ.get('NEXAAPI_KEY', 'YOUR_API_KEY'))

def agent_image_task(prompt: str) -> str:
    """AI agent action: generate image from prompt"""
    response = client.image.generate(
        model='flux-schnell',
        prompt=prompt,
        width=1024,
        height=1024
    )
    return response.image_url

# Usage in agent workflow
image_url = agent_image_task("a professional product photo of a smartphone")
print(f"Generated: {image_url}")
print("Cost: $0.003")

from nexaapi import NexaAPI
import os

client = NexaAPI(api_key=os.environ.get('NEXAAPI_KEY'))

class MultimodalAgent:
    """AI agent with multimodal capabilities powered by NexaAPI"""
    
    def __init__(self):
        self.client = client
        self.task_log = []
    
    def generate_report_image(self, data_summary: str) -> str:
        """Generate a visual report image"""
        prompt = f"Professional data visualization chart showing: {data_summary}"
        result = self.client.image.generate(
            model='flux-schnell',
            prompt=prompt
        )
        self.task_log.append(f"Generated image: {result.image_url}")
        return result.image_url
    
    def narrate_summary(self, text: str) -> str:
        """Convert text summary to audio"""
        result = self.client.audio.tts(text=text, voice='alloy')
        output_path = '/tmp/agent_report.mp3'
        with open(output_path, 'wb') as f:
            f.write(result.audio_data)
        self.task_log.append(f"Generated audio: {output_path}")
        return output_path
    
    def run_workflow(self, task: str) -> dict:
        """Execute a complete multimodal agent workflow"""
        print(f"🤖 Agent executing: {task}")
        
        # Generate visual
        image_url = self.generate_report_image(task)
        
        # Generate audio summary
        audio_path = self.narrate_summary(f"Task completed: {task}")
        
        return {
            "task": task,
            "image": image_url,
            "audio": audio_path,
            "cost": "$0.003",
            "log": self.task_log
        }

# Run the agent
agent = MultimodalAgent()
result = agent.run_workflow("quarterly sales performance analysis")
print(result)

// npm install nexaapi
import NexaAPI from 'nexaapi';

const client = new NexaAPI({ apiKey: process.env.NEXAAPI_KEY });

async function runAgentWorkflow(task) {
  console.log(`🤖 Agent executing: ${task}`);
  
  // Generate image
  const imageResult = await client.image.generate({
    model: 'flux-schnell',
    prompt: `Professional visualization of: ${task}`
  });
  
  // Generate audio
  const audioResult = await client.audio.tts({
    text: `Task completed: ${task}`,
    voice: 'alloy'
  });
  
  return {
    task,
    image: imageResult.imageUrl,
    cost: '$0.003'
  };
}

runAgentWorkflow('quarterly sales analysis').then(console.log);