AI Field Automator

The AI Field Automator module lets you attach AI generation rules directly to Backdrop field instances. When enabled on a field, the module can populate that field from another field, from rendered entity content, from tokens, from media inputs, or from search results depending on the selected automator type.

This module is intended for the Backdrop ai ecosystem and follows the same provider/model configuration used by the rest of the AI modules in this codebase.

• Per-field AI automator configuration in Field UI
• Multiple worker modes: direct, batch, queue/cron, and widget action
• Rule ordering by weight
• Text, scalar, reference, media, structured, search, and domain-specific automator types
• Optional async widget-action generation when ai_async is enabled

• Backdrop CMS 1.x
• PHP 8.2 or later
• The ai module

Some automator types require additional modules or field types, for example:

• token support for token-mode prompts
• ai_search for vector-search automators
• media/file/image-related modules for media automators

1. Place the module in modules/custom/
2. Enable it with bee pm-enable ai_field_automator or from admin/modules
3. Make sure at least one AI provider/model is configured

1. Go to a field settings form, for example: admin/structure/types/manage/[content-type]/fields/[field]/edit
2. Open the AI Automator fieldset
3. Enable the automator for that field
4. Choose an automator type
5. Configure the prompt, base field, model, worker type, and any type-specific settings
6. Save the field

The module also provides an overview page at:

• /admin/config/ai/ai-field-automator

• Direct runs during save
• Batch runs via Backdrop batch processing after form submit
• Queue/Cron queues work for cron
• Widget action adds a Generate with AI button to supported field widgets

The module includes a broad set of automator types for:

• Text fields
• Numeric, boolean, email, telephone, link, and option fields
• Taxonomy and entity references
• Images, audio, speech, and video workflows
• JSON, metatag, moderation-state, address, office-hours, FAQ, and chart data
• Search API AI vector search and Views extraction

The full type reference is documented in:

• AUTOMATOR_TYPES.md

Implementation details and module architecture are documented separately in:

• HOW_IT_WORKS.md

Focused Xdebug notes for the post.field_summary widget-action automator are in:

• DEBUGGING_FIELD_SUMMARY.md

• Configuration is stored per field, not as standalone config entities
• The module creates an ai_automator_status field on bundles that have active automators
• For queue/cron workers, choose an explicit text format when needed because cron may run as anonymous

Bugs and feature requests should be reported in the Issue Queue.

Justin Keiser

• Created for Backdrop CMS by Justin Keiser.
• Inspired by the Drupal AI Automators module.
• Developed with AI assistance.

This project is GPL v2 software. See the LICENSE.txt file in this directory for complete text.