Ask My Content - AI Search Chatbot

On the Payments tab you can see your tokens usage, balance, and make the payments. You get one-time US $5.00 credit, which is applied to third-party AI services (like OpenAI and Gemini) indexing/embedding and chat completion/answers.

The admin "Start Indexing" button schedules a background task via WordPress cron. Caching, blocked loopback requests, or a missing server cron runner can prevent that task from starting. The settings page checks queued jobs while it is open, retries cron dispatch at most once every 30 seconds, and shows a warning if indexing remains scheduled for more than two minutes. Solutions:

1. Use Stop Indexing, then Start Indexing to retry. Scheduling failures and missing workers now display an error instead of remaining scheduled indefinitely.
2. Ask your hosting provider to check WordPress cron and loopback requests. If automatic WP-Cron is disabled, a server cron runner must execute queued jobs.
3. If your host provides WP-CLI, use wp amc index from your WordPress directory. It runs synchronously and uses your saved content-type settings. See Command Line Usage. For detailed diagnostics on sites you manage, enable ASKMYCO_DEBUG, WP_DEBUG, and WP_DEBUG_LOG in wp-config.php, with WP_DEBUG_DISPLAY disabled. AMC cron log entries include the site URL, hook, scheduled time, dispatch attempts, lock contention, worker start, and completion. Repeated status polls do not repeat the same warning.

Enable Automatically re-index modified pages, posts, and content if you want content edits to appear in chat results quickly without running manual indexing. Keep it disabled if your site has frequent edits to large pages (for example long landing pages or content-heavy builder pages). In that case, each update can trigger additional background embedding requests, which may increase API usage and add server/backend load. Practical guideline:

• Enable it for low-to-moderate edit frequency or smaller pages.
• Disable it for high-frequency editing of large pages, and run indexing in batches when convenient.

Low indexed token count usually means the plugin did not find much readable page content to send for indexing. This can happen if:

1. The page is mostly built from theme templates, banners, sliders, or other repeated global sections.
2. The important content is loaded by a plugin or builder widget that does not expose much text to WordPress.
3. Indexing finished before the latest content changes were re-run. Ask My Content now supports common WordPress content builders much better, including Gutenberg, Elementor, Avada, and WooCommerce product content.

Ask My Content is designed to index the main readable text of your site content. Currently supported:

• Standard WordPress pages and posts
• Gutenberg block content
• Elementor content in many common setups
• Avada content in many common setups
• WooCommerce product content, including product descriptions, short descriptions, price, SKU, weight, dimensions, and attributes
• Other public post types, if enabled in plugin settings
• Server-rendered homepage sections generated by themes and the Customizer, indexed separately from the normal front-page content In general, the plugin reads the main post content the way WordPress normally renders it, using the standard content pipeline (apply_filters('the_content', ...)).

• headers or footers
• sliders, popups, and announcement bars
• heavily customized builder widgets
• content injected from external sources or loaded later with JavaScript If a page is important for search quality, open it in WordPress and make sure the main text is actually stored in the page, post, or product content itself before re-running indexing.

By default, the plugin uses the developer's AI API key on a secure backend (with generous usage limits). This means you don't need your own AI API key.

No. Heavy AI processing (embeddings, search, completions) happens on a separate Node.js backend service.

Ask My Content indexes publicly accessible website content, including selected pages, posts, products, and other public content types. Indexed content is securely transmitted to the Ask My Content cloud backend, where text and vector embeddings are stored to enable semantic search and AI-generated answers. Relevant content excerpts may also be transmitted to third-party AI providers for processing. Content is synchronized when updated, and deleted posts are automatically removed from the search index.

When visitors use the chatbot, their questions are transmitted to the Ask My Content backend and may be processed by third-party AI providers. Questions and generated answers may be stored in chat history and made available to the website administrator. Website administrators should inform visitors about this processing in their website's privacy policy and avoid requesting sensitive personal information through the chatbot.

Communication between the WordPress plugin and the Ask My Content backend uses HTTPS encryption. Website administrators should ensure that content selected for indexing is appropriate for processing by external cloud services.

Complete question-and-answer histories are stored locally in the website's WordPress database and are accessible to the website administrator. The Ask My Content cloud backend does not maintain a centralized database of complete chat histories. Limited diagnostic information may be retained in server logs for approximately 30 days.

Not yet. Ask My Content is not currently compatible with WordPress multisite installations, so it must be activated on individual sites (no network-wide support yet).

Yes. The [ask_my_content] shortcode accepts optional attributes such as title, sessionId, width, height, and the various color tokens (headerBackground, headerTextColor, chatBackground, chatUserBubble, chatBotBubble, inputBackground, inputTextColor). They map to the same style variables used by the block. See the inline comment above amc_render_chat_shortcode() in ask-my-content.php for the current list and defaults.