Jobs Endpoints

Last updated: March 17, 2026

Implementation: inc/Api/Jobs.php

Base URL: /wp-json/datamachine/v1/jobs

Overview

Jobs endpoints provide monitoring and management of workflow executions. Jobs represent individual execution instances of flows, including batch job hierarchies where a parent job spawns child jobs for parallel processing.

Authentication

Requires manage_flows permission (checked via PermissionHelper). Supports user-scoped and agent-scoped filtering for multi-agent environments.

React Interface

The Jobs interface is a React-based management dashboard built on @wordpress/components and TanStack Query.

Endpoints

GET /jobs

Retrieve jobs with filtering, sorting, and pagination.

Permission: manage_flows capability required

Parameters:

orderby (string, optional): Order jobs by field (default: job_id)
order (string, optional): Sort order – ASC or DESC (default: DESC)
per_page (integer, optional): Number of jobs per page (default: 50, max: 100)
offset (integer, optional): Offset for pagination (default: 0)
pipeline_id (integer, optional): Filter by pipeline ID
flow_id (integer, optional): Filter by flow ID
status (string, optional): Filter by job status (completed, failed, processing, etc.)
user_id (integer, optional): Filter by user ID (admin only; non-admins always see own data)

Example Requests:

bash

# Get all jobs (recent first)
curl https://example.com/wp-json/datamachine/v1/jobs 
  -u username:application_password

# Get failed jobs only
curl https://example.com/wp-json/datamachine/v1/jobs?status=failed 
  -u username:application_password

# Get jobs for specific flow with pagination
curl https://example.com/wp-json/datamachine/v1/jobs?flow_id=42&per_page=25&offset=0 
  -u username:application_password

# Get jobs for specific pipeline
curl https://example.com/wp-json/datamachine/v1/jobs?pipeline_id=5 
  -u username:application_password

Success Response (200 OK):

json

{
  "success": true,
  "data": [
    {
      "job_id": 1523,
      "flow_id": 42,
      "pipeline_id": 5,
      "status": "completed",
      "parent_job_id": null,
      "source": "scheduler",
      "started_at": "2024-01-02 14:30:00",
      "completed_at": "2024-01-02 14:30:15",
      "error_message": null
    },
    {
      "job_id": 1522,
      "flow_id": 42,
      "pipeline_id": 5,
      "status": "failed",
      "parent_job_id": null,
      "source": "manual",
      "started_at": "2024-01-02 14:00:00",
      "completed_at": "2024-01-02 14:00:05",
      "error_message": "Handler configuration missing"
    }
  ],
  "total": 1523,
  "per_page": 50,
  "offset": 0
}

Response Fields:

success (boolean): Request success status
data (array): Array of job objects
total (integer): Total number of jobs matching filters
per_page (integer): Number of jobs per page
offset (integer): Pagination offset

Job Object Fields:

job_id (integer): Unique job identifier
flow_id (integer): Associated flow ID
pipeline_id (integer): Associated pipeline ID
status (string): Job status (see statuses below)
parent_job_id (integer|null): Parent job ID for batch child jobs (null for top-level jobs)
source (string): Job trigger source (scheduler, manual, system, pipeline_system_task, chat)
started_at (string): Job start timestamp
completed_at (string|null): Job completion timestamp
error_message (string|null): Error message if failed
engine_data (object): Job execution data including task parameters and effects (for system tasks)

Job Statuses:

pending – Job queued but not started
processing – Currently executing
completed – Successfully completed
completed_no_items – Completed successfully but no new items found to process
agent_skipped – Completed intentionally without processing the current item (supports compound statuses like agent_skipped - {reason})
failed – Execution failed with error

GET /jobs/{id}

Get a specific job by ID with full details.

Permission: manage_flows capability required

Parameters:

id (integer, required): Job ID

Example Request:

bash

curl https://example.com/wp-json/datamachine/v1/jobs/1523 
  -u username:application_password

Success Response (200 OK):

json

{
  "success": true,
  "data": {
    "job_id": 1523,
    "flow_id": 42,
    "pipeline_id": 5,
    "status": "completed",
    "parent_job_id": null,
    "source": "scheduler",
    "started_at": "2024-01-02 14:30:00",
    "completed_at": "2024-01-02 14:30:15",
    "error_message": null,
    "engine_data": {}
  }
}

Error Response (404 Not Found):

json

{
  "code": "job_not_found",
  "message": "Job not found.",
  "data": {"status": 404}
}

DELETE /jobs

Clear jobs from the database.

Permission: manage_flows capability required

Parameters:

type (string, required): Which jobs to clear – all or failed
cleanup_processed (boolean, optional): Also clear processed items tracking (default: false)

Example Requests:

bash

# Clear all jobs
curl -X DELETE https://example.com/wp-json/datamachine/v1/jobs 
  -H "Content-Type: application/json" 
  -u username:application_password 
  -d '{"type": "all"}'

# Clear failed jobs only
curl -X DELETE https://example.com/wp-json/datamachine/v1/jobs 
  -H "Content-Type: application/json" 
  -u username:application_password 
  -d '{"type": "failed"}'

# Clear all jobs and processed items
curl -X DELETE https://example.com/wp-json/datamachine/v1/jobs 
  -H "Content-Type: application/json" 
  -u username:application_password 
  -d '{"type": "all", "cleanup_processed": true}'

Success Response (200 OK):

json

{
  "success": true,
  "message": "Jobs cleared successfully.",
  "jobs_deleted": 42,
  "processed_items_cleaned": false
}

Error Response (400 Bad Request):

json

{
  "code": "invalid_type",
  "message": "Invalid type parameter. Must be 'all' or 'failed'.",
  "data": {"status": 400}
}

Batch Job Hierarchy

Jobs support parent-child relationships for batch processing. When a pipeline uses the PipelineBatchScheduler, a parent job spawns multiple child jobs that execute in parallel.

Structure

Parent Job (job_id: 100, parent_job_id: null)
├── Child Job (job_id: 101, parent_job_id: 100)
├── Child Job (job_id: 102, parent_job_id: 100)
└── Child Job (job_id: 103, parent_job_id: 100)

Key behaviors:

Parent job status remains processing until all children complete
parent_job_id field links children to their parent
Child completion triggers PipelineBatchScheduler::onChildComplete() to check if all children are done
Batch size and chunk processing are configured at the pipeline level

Querying Batch Jobs

bash

# Get children of a batch parent
curl "https://example.com/wp-json/datamachine/v1/jobs?parent_job_id=100" 
  -u username:application_password

Job Undo System

Implementation: inc/Engine/AI/System/Tasks/SystemTask.php (@since v0.33.0)

System tasks that modify WordPress content can opt into undo support. The undo system reads a standardized effects array from the job’s engine_data and reverses each effect in reverse order.

Supported Undo Effect Types

Effect Type	Reversal Action
`post_content_modified`	Restores from WordPress revision
`post_meta_set`	Restores previous value or deletes meta
`attachment_created`	Deletes the attachment
`featured_image_set`	Removes or restores previous thumbnail

How Undo Works

During execution, tasks record effects in engine_data.effects:

json

{
  "effects": [
    {
      "type": "post_content_modified",
      "post_id": 42,
      "revision_id": 123
    },
    {
      "type": "post_meta_set",
      "post_id": 42,
      "meta_key": "description",
      "previous_value": "old value"
    }
  ]
}

Tasks opt in by returning true from supportsUndo()
Undo reverses effects in reverse order (last effect first)
Unknown effect types are skipped (not failed) so tasks with mixed reversible/irreversible effects degrade gracefully

CLI Undo

json

{
  "effects": [
    {
      "type": "post_content_modified",
      "post_id": 42,
      "revision_id": 123
    },
    {
      "type": "post_meta_set",
      "post_id": 42,
      "meta_key": "description",
      "previous_value": "old value"
    }
  ]
}

Undo Response Format

json

{
  "effects": [
    {
      "type": "post_content_modified",
      "post_id": 42,
      "revision_id": 123
    },
    {
      "type": "post_meta_set",
      "post_id": 42,
      "meta_key": "description",
      "previous_value": "old value"
    }
  ]
}

Common Workflows

Monitor Flow Execution

bash

# Undo a job's effects
wp datamachine jobs undo <job_id> --allow-root

# Dry run to preview what would be reverted
wp datamachine jobs undo <job_id> --dry-run --allow-root

# Undo specific task type only
wp datamachine jobs undo <job_id> --task-type=alt_text_generation --allow-root

# Force undo even on non-completed jobs
wp datamachine jobs undo <job_id> --force --allow-root

Debug Failed Executions

json

{
  "success": true,
  "reverted": [
    {
      "type": "post_content_modified",
      "post_id": 42,
      "message": "Restored from revision 123"
    }
  ],
  "skipped": [],
  "failed": []
}

Cleanup Job History

bash

# Get recent jobs for specific flow
curl https://example.com/wp-json/datamachine/v1/jobs?flow_id=42&per_page=10 
  -u username:application_password

Integration Examples

Python Job Monitoring

bash

# Get all failed jobs
curl https://example.com/wp-json/datamachine/v1/jobs?status=failed 
  -u username:application_password

JavaScript Job Dashboard

bash

# Clear failed jobs to reset execution history
curl -X DELETE https://example.com/wp-json/datamachine/v1/jobs 
  -H "Content-Type: application/json" 
  -u username:application_password 
  -d '{"type": "failed"}'

PHP Job Monitoring

python

import requests
from requests.auth import HTTPBasicAuth

url = "https://example.com/wp-json/datamachine/v1/jobs"
auth = HTTPBasicAuth("username", "application_password")

# Get failed jobs
params = {"status": "failed", "per_page": 100}
response = requests.get(url, params=params, auth=auth)

if response.status_code == 200:
    data = response.json()
    print(f"Found {len(data['data'])} failed jobs")

    for job in data['data']: 
        print(f"Job {job['job_id']}: {job['error_message']}")
else:
    print(f"Error: {response.json()['message']}")

Use Cases

Execution Monitoring

During execution, tasks record effects in engine_data.effects:

javascript

const axios = require('axios');

const jobAPI = {
  baseURL: 'https://example.com/wp-json/datamachine/v1/jobs',
  auth: {
    username: 'admin',
    password: 'application_password'
  }
};

// Get job statistics
async function getJobStats(flowId) {
  const params = { flow_id: flowId, per_page: 100 };
  const response = await axios.get(jobAPI.baseURL, {
    params,
    auth: jobAPI.auth
  });

  const jobs = response.data.data;
  const stats = {
    total: jobs.length,
    completed: jobs.filter(j => j.status === 'completed').length,
    failed: jobs.filter(j => j.status === 'failed').length,
    processing: jobs.filter(j => j.status === 'processing').length
  };

  return stats;
}

// Clear old jobs
async function clearJobs(type = 'failed') {
  const response = await axios.delete(jobAPI.baseURL, {
    data: { type },
    auth: jobAPI.auth
  });

  return response.data.success;
}

// Usage
const stats = await getJobStats(42);
console.log(`Flow 42: ${stats.completed} completed, ${stats.failed} failed`);

await clearJobs('failed');
console.log('Failed jobs cleared');

Performance Analysis

Tasks opt in by returning true from supportsUndo()

php

$url = 'https://example.com/wp-json/datamachine/v1/jobs';
$auth = base64_encode('username:application_password');

// Get jobs for specific flow
$params = http_build_query([
    'flow_id' => 42,
    'status' => 'failed',
    'per_page' => 50
]);

$response = wp_remote_get($url . '?' . $params, [
    'headers' => [
        'Authorization' => 'Basic ' . $auth
    ]
]);

if (!is_wp_error($response)) {
    $data = json_decode(wp_remote_retrieve_body($response), true);

    foreach ($data['data'] as $job) {
        error_log(sprintf(
            'Job %d failed: %s',
            $job['job_id'],
            $job['error_message']
        ));
    }
}

Cleanup and Maintenance

Undo reverses effects in reverse order (last effect first)

bash

# Get all jobs ordered by completion time
curl https://example.com/wp-json/datamachine/v1/jobs?orderby=completed_at&order=DESC 
  -u username:application_password

Execute – Flow execution
Flows – Flow management
Processed Items – Deduplication tracking
Logs – Detailed execution logs

Unknown effect types are skipped (not failed) so tasks with mixed reversible/irreversible effects degrade gracefully

Jobs Endpoints

Overview

Authentication

React Interface

Endpoints

GET /jobs

GET /jobs/{id}

DELETE /jobs

Batch Job Hierarchy

Structure

Querying Batch Jobs

Job Undo System

Supported Undo Effect Types

How Undo Works

CLI Undo

Undo Response Format

Common Workflows

Monitor Flow Execution

Debug Failed Executions

Cleanup Job History

Integration Examples

Python Job Monitoring

JavaScript Job Dashboard

PHP Job Monitoring

Use Cases

Execution Monitoring

Performance Analysis

Cleanup and Maintenance

Related Documentation

More in Endpoints

Chat Endpoint

Users Endpoints

Error Handling

Analytics Endpoints