This package provides a Go client implementation for the Agent-to-Agent (A2A) communication protocol.

• JSON-RPC 2.0 compliant client
• Supports core A2A methods:
  • message/send: Send a new task
  • tasks/get: Get task status
  • tasks/cancel: Cancel a task
• Streaming task updates with Server-Sent Events (SSE)
• Error handling with A2A error codes
• Type-safe request/response handling

package main

import (
    "log"
    "a2a/client"
    "a2a/models"
)

func main() {
    // Create a new client
    a2aClient := client.NewClient("http://localhost:8080")

    // Create a task message
    message := models.Message{
        Role: "user",
        Parts: []models.Part{
            {
                Type: stringPtr("text"),
                Text: stringPtr("Hello, A2A agent!"),
            },
        },
    }

    // Send a task
    response, err := a2aClient.SendTask(models.TaskSendParams{
        ID:      "task-1",
        Message: message,
    })
    if err != nil {
        log.Fatalf("Failed to send task: %v", err)
    }

    // Get the task from the response
    task, ok := response.Result.(*models.Task)
    if !ok {
        log.Fatalf("Expected result to be a Task")
    }

    // Use the task...
}

func NewClient(baseURL string) *Client

Creates a new A2A client instance with the specified base URL.

func (c *Client) SendTask(params models.TaskSendParams) (*models.JSONRPCResponse, error)

Sends a new task to the agent. Returns a JSON-RPC response containing the task or an error.

func (c *Client) GetTask(params models.TaskQueryParams) (*models.JSONRPCResponse, error)

Gets the status of a task. Returns a JSON-RPC response containing the task or an error.

func (c *Client) CancelTask(params models.TaskIDParams) (*models.JSONRPCResponse, error)

Cancels a task. Returns a JSON-RPC response containing the task or an error.

The client supports streaming task updates using Server-Sent Events (SSE). To use streaming:

1. Set the Accept header to text/event-stream in your request
2. The server will respond with a stream of task status updates
3. Each update will be a JSON object containing:
   • Task ID
   • Current status
   • Whether it's the final update

Example streaming usage:

// Create a task with streaming
message := models.Message{
    Role: "user",
    Parts: []models.Part{
        {
            Type: stringPtr("text"),
            Text: stringPtr("Hello, A2A agent!"),
        },
    },
}

// Send a task with streaming enabled
response, err := a2aClient.SendTaskWithStreaming(models.TaskSendParams{
    ID:      "task-1",
    Message: message,
})
if err != nil {
    log.Fatalf("Failed to send task: %v", err)
}

// Process streaming updates
for update := range response.Updates {
    if update.Error != nil {
        log.Printf("Error: %v", update.Error)
        continue
    }
    
    // Process the update
    log.Printf("Task %s: %s", update.Result.ID, update.Result.Status.State)
}

Run the tests with:

go test ./...

The test suite includes examples of:

• Sending tasks
• Getting task status
• Canceling tasks
• Streaming task updates
• Error handling

Important: The sample code provided is for demonstration purposes and illustrates the mechanics of the Agent-to-Agent (A2A) protocol. When building production applications, it is critical to treat any agent operating outside of your direct control as a potentially untrusted entity.

All data received from an external agent—including but not limited to its AgentCard, messages, artifacts, and task statuses—should be handled as untrusted input. For example, a malicious agent could provide an AgentCard containing crafted data in its fields (e.g., description, name, skills.description). If this data is used without sanitization to construct prompts for a Large Language Model (LLM), it could expose your application to prompt injection attacks. Failure to properly validate and sanitize this data before use can introduce security vulnerabilities into your application.

Developers are responsible for implementing appropriate security measures, such as input validation and secure handling of credentials to protect their systems and users.