​

RunAgent Go Client Example

This example demonstrates how to use the RunAgent Go client to connect to and execute an agent for research tasks.

​

Prerequisites

• Go 1.18 or later
• RunAgent server running locally on port 8453
• Access to the research_crew agent

​

Installation

First, install the RunAgent Go client:

go mod init your-project
go get github.com/runagent-dev/runagent-go/pkg/client

​

Basic Usage

package main

import (
	"context"
	"fmt"
	"log"
	"time"

	"github.com/runagent-dev/runagent-go/pkg/client"
)

func main() {
	// Create a new client connection
	agentClient, err := client.NewWithAddress(
		"331fa66f-1089-43f0-8912-cc1b4c22ab01", // Agent ID
		"research_crew",                        // Agent name
		true,                                   // Enable SSL/TLS
		"localhost",                            // Server address
		8453,                                   // Server port
	)
	if err != nil {
		log.Fatalf("Failed to create client: %v", err)
	}
	defer agentClient.Close()

	// Create a context with timeout
	ctx, cancel := context.WithTimeout(context.Background(), 2*time.Minute)
	defer cancel()

	// Execute the agent with parameters
	result, err := agentClient.Run(ctx, map[string]interface{}{
		"topic": "global warming",
	})
	if err != nil {
		log.Fatalf("Failed to run agent: %v", err)
	}

	fmt.Printf("Result: %v\n", result)
}

​

Code Breakdown

​

Client Initialization

The client.NewWithAddress() function creates a new RunAgent client with the following parameters:

• Agent ID: "331fa66f-1089-43f0-8912-cc1b4c22ab01" - Unique identifier for the agent
• Agent Name: "research_crew" - Human-readable name of the agent
• SSL/TLS: true - Enable secure connection
• Address: "localhost" - Server hostname or IP address
• Port: 8453 - Server port number

​

Context and Timeout

ctx, cancel := context.WithTimeout(context.Background(), 2*time.Minute)
defer cancel()

A context with a 2-minute timeout is created to prevent the agent execution from running indefinitely. The defer cancel() ensures proper cleanup.

​

Agent Execution

result, err := agentClient.Run(ctx, map[string]interface{}{
	"topic": "global warming",
})

The Run() method executes the agent with the provided parameters. In this example, we’re passing a topic parameter with the value "global warming".

​

Configuration Options

You can customize the client behavior by modifying the connection parameters:

// For remote server
agentClient, err := client.NewWithAddress(
	"your-agent-id",
	"your-agent-name",
	true,           // Use HTTPS
	"agent.example.com", // Remote server
	443,            // HTTPS port
)

// For local development without SSL
agentClient, err := client.NewWithAddress(
	"your-agent-id",
	"your-agent-name",
	false,          // Disable SSL
	"localhost",
	8080,           // HTTP port
)

​

Error Handling

The example includes basic error handling for:

• Client creation failures
• Agent execution failures
• Context timeouts

if err != nil {
	log.Fatalf("Failed to create client: %v", err)
}

​

Best Practices

1. Always close the client: Use defer agentClient.Close() to ensure proper cleanup
2. Set appropriate timeouts: Use context with timeout to prevent hanging operations
3. Handle errors gracefully: Check for errors at each step
4. Secure connections: Use SSL/TLS for production environments

​

Common Parameters

Different agents may accept different parameters. Common examples include:

// Research agent
result, err := agentClient.Run(ctx, map[string]interface{}{
	"topic": "artificial intelligence",
	"depth": "detailed",
	"sources": []string{"academic", "news"},
})

// Data processing agent
result, err := agentClient.Run(ctx, map[string]interface{}{
	"input_file": "/path/to/data.csv",
	"operation": "analyze",
	"format": "json",
})

​

Troubleshooting

• Connection refused: Ensure the RunAgent server is running on the specified port
• Timeout errors: Increase the context timeout for long-running operations
• SSL/TLS errors: Verify the SSL configuration matches your server setup
• Agent not found: Check that the agent ID and name are correct

​

Next Steps

• Explore different agent types and their parameters
• Implement more sophisticated error handling
• Add logging and monitoring
• Consider using connection pooling for multiple requests