> ## Documentation Index
> Fetch the complete documentation index at: https://docs.cloudcoding.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Parallel Execution

> Run multiple coding tasks concurrently with the /invoke/stdout endpoint.

## Overview

The `/invoke/stdout` endpoint supports parallel execution, allowing you to run multiple coding tasks concurrently for faster development workflows. Each task runs independently, and you can track its output in real-time.

## Request Structure

To run tasks in parallel, provide a `parallel_tasks` array in your request to the `/invoke/stdout` endpoint. See the [Quickstart](/quickstart) for details on authentication and other request parameters.

```json theme={null}
{
  "repo_url": "https://github.com/username/repo",
  "github_token": "YOUR_GITHUB_TOKEN",
  "branch": "main",
  "parallel_tasks": [
    {
      "prompt": "Create comprehensive API documentation",
      "task_name": "api-docs"
    },
    {
      "prompt": "Add unit tests for all components",
      "task_name": "unit-tests"
    }
  ]
}
```

## Task Configuration

Each object in the `parallel_tasks` array requires the following fields:

| Field       | Type   | Required | Description                       |
| ----------- | ------ | -------- | --------------------------------- |
| `prompt`    | string | ✅        | The prompt/query for this task.   |
| `task_name` | string | ✅        | A unique identifier for the task. |

## Example

This example runs two tasks in parallel. The stream will contain interleaved `custom` events from each task, identified by `task_name`.

```bash theme={null}
curl -X POST "https://api.cloudcoding.ai/invoke/stdout" \
  -H "Content-Type: application/json" \
  -H "x-api-key: YOUR_API_KEY" \
  --no-buffer \
  -d '{
    "repo_url": "https://github.com/username/my-app",
    "github_token": "YOUR_GITHUB_TOKEN",
    "parallel_tasks": [
      {
        "prompt": "Add comprehensive error handling",
        "task_name": "error-handling"
      },
      {
        "prompt": "Create unit tests for all components",
        "task_name": "testing"
      }
    ]
  }'
```

## Response Stream

The API returns a stream of Server-Sent Events (SSE). You can distinguish between outputs from different parallel tasks by checking the `task_name` in the `data` payload of `custom` events.

**Example `custom` event from the "testing" task:**

```json theme={null}
{
  "event": "custom",
  "data": {
    "task_name": "testing",
    "stdout": "Running tests for component A...\n"
  }
}
```

Refer to the [Quickstart](/quickstart) for more details on the response stream.

## Next Steps

<CardGroup>
  <Card title="Quickstart" icon="rocket" href="/quickstart">
    Get started with the API and single task invocations
  </Card>

  <Card title="Dashboard" icon="chart-line" href="https://cloudcoding.ai/home">
    Manage your API keys and monitor usage
  </Card>
</CardGroup>