ai_accessibility

Detect accessibility issues and get automatic suggestions powered by ChatGPT and other AIs to improve inclusiveness in your Drupal site.

Features

• Scans content forms using axe-core to detect common accessibility issues.
• Integrates ChatGPT to provide automatic suggestions (e.g., alt text for images).
• Designed as a multi-AI framework: future versions will allow connecting multiple AI providers (e.g., OpenAI, Hugging Face, Anthropic).
• Displays an on-screen interactive panel with detected violations and possible fixes.
• Provides simple endpoints (mocked in MVP) for validation and suggestions.

Use cases:

• Editors who want to ensure accessibility without relying on technical experts.
• Sites aiming for faster compliance with WCAG guidelines.
• Teams experimenting with AI to enhance accessibility in their editorial workflow.

Post-Installation

1. Place the module in modules/ai_accessibility.
2. Enable it with drush en ai_accessibility or through the module UI.
3. Configure basic settings at: /admin/config/content/ai-accessibility
4. When editing a node, the JS library is injected and an accessibility + AI suggestions panel will appear.

Additional Requirements

• Drupal Core 10 or 11.
• This MVP uses mocked endpoints, so no external API is required yet.
• To use the ChatGPT integration, an OpenAI API key will be required (future versions will support other providers as well).

Recommended modules/libraries

• CKEditor Accessibility Checker → in-editor accessibility analysis.
• Automatic Alternative Text → generates alt text using external APIs.

Similar projects

• CKEditor Accessibility Checker: focuses only on the editor, without AI support.
• Automatic Alternative Text: generates alt text but does not validate other accessibility issues.
• AI Accessibility: combines validation (axe-core) + AI suggestions, starting with ChatGPT but designed for multiple AI providers.

Supporting this Module

This module is experimental (MVP stage). You can support it by reporting issues, contributing patches, or helping shape the roadmap.

Community Documentation

• Demos and walkthroughs will be shared soon (DrupalPod / YouTube).
• The issue queue will serve as the main space to document progress and discuss new AI integrations.