WordPress Plugin for AI-Generated SEO Meta Descriptions

https://github.com/woodchen-ink/wordpress-auto-meta-ai

Project Overview

wordpress-auto-meta-ai is a lightweight WordPress plugin for content creators and site operators, designed to give every post a more complete SEO meta description without adding editorial work. As publishing picks up, meta descriptions are often missing or inconsistent, affecting search snippets and click-through rates. This project uses AI through OpenAI-compatible APIs to generate suitable descriptions when posts are first published, skipping existing meta descriptions by default to avoid accidental overwrites. Alongside automation, it offers bulk generation from the post list with real-time progress updates, making it easy to fill gaps in older content. Results can be saved to Yoast SEO, Rank Math, or native WordPress excerpts as needed, reducing the cost of switching between plugin ecosystems. Custom prompt templates and support for multiple compatible APIs, including local models, give different sites a controlled way to produce consistent SEO descriptions.

Project Statistics

📊 Project at a Glance

⭐ 0 Stars | 🍴 0 Forks | 👀 0 Watchers | 🐛 0 Issues

📅 Created: 2026-03-03 | 🔄 Updated: 2026-03-03

💻 Tech Stack

The project is developed primarily in PHP, which accounts for 100.0% of the codebase

Key Features

• Generate on publication, with less effort: Automatically triggers AI generation of SEO meta descriptions when a post is first published, removing the need to agonize over what to write and streamlining publication.
• One-click bulk processing: Select multiple posts in the post list to generate meta descriptions in bulk, with real-time progress feedback—ideal for reviewing and filling gaps in older posts' SEO metadata.
• Save to multiple destinations: Save generated results to the meta description fields in Yoast SEO or Rank Math, or to native WordPress excerpts (post_excerpt), so one result can serve multiple destinations.
• Skip existing descriptions by default: Automatically skips posts that already have meta descriptions, protecting carefully edited content from accidental overwrites.
• OpenAI-compatible APIs for greater flexibility: Supports not only OpenAI but also compatible services such as DeepSeek, LiteLLM, and Ollama, letting you switch based on cost, speed, and self-hosting requirements.
• Custom prompts to match your style: Edit prompt templates directly in the settings page and use {post_content} to insert post content, making it easy to control output length, tone, and keyword density. More features are being refined.

Technical Architecture

The plugin uses a lightweight WordPress + PHP 7.4+ architecture, designed around generating descriptions on publication and filling gaps in bulk. It relies on WordPress Hooks: it listens for a post's first publication to trigger generation, and provides a bulk action in the admin post list. Asynchronous requests drive queue-style processing, while a real-time progress bar reports execution status, reducing the risk of timeouts from processing everything at once. AI integration uses an OpenAI Chat Completions-compatible API, with configurable API URLs, models, and keys. Implementations such as DeepSeek and LiteLLM/Ollama can be substituted, decoupling the plugin from providers through API compatibility. The default generation policy is not to overwrite existing data: it checks target fields before writing to avoid disrupting existing SEO settings. A multi-destination saving approach allows the same result to be written to Yoast (_yoast_wpseo_metadesc), Rank Math (rank_math_description), and native WordPress excerpts (post_excerpt), broadening compatibility across SEO ecosystems. Prompt templates can be maintained in the settings page, using the {post_content} placeholder to insert the post body for controllable, iterative improvements to generation quality. The workflow is: trigger event → read post content → render prompt → call compatible API → parse result → write to selected destinations and record status.

Usage Guide

• Installation: Download the latest .zip from Releases. In the admin dashboard, go to Plugins → Add New Plugin → Upload Plugin, select the file, and click “Install Now,” then activate it. For manual installation, extract it to wp-content/plugins/wordpress-auto-meta-ai/ and activate it from the Plugins page. You can also install it from the command line: wp plugin install https://github.com/woodchen-ink/wordpress-auto-meta-ai/releases/latest/download/wordpress-auto-meta-ai.zip --activate
• Configuration: Go to Settings → Auto Meta AI, enter your API details, and save. Suggested configuration:
  • API endpoint: https://api.openai.com/v1/chat/completions (use the corresponding endpoint for DeepSeek/LiteLLM/Ollama)
  • Model: gpt-3.5-turbo (or your compatible model name)
  • Prompt template (example): 请基于以下内容生成不超过160字的中文SEO元描述，避免引号与换行：{post_content}
  • Save destinations: Select Yoast SEO (_yoast_wpseo_metadesc), Rank Math (rank_math_description), and/or WordPress excerpts (post_excerpt)
• Automatic generation: Generation is triggered when a post is first published. If a target destination already contains a description, it is skipped by default rather than overwritten.
• Bulk generation: Under Posts → All Posts, select multiple posts, choose the bulk action to generate meta descriptions, and run it. The page will display real-time progress. Test the model and prompt on a few posts before processing in bulk.

Summary and Assessment

This plugin addresses the common need to generate SEO meta descriptions on publication. It covers both automatic generation and bulk generation in the admin dashboard, can write to Yoast, Rank Math, and native excerpts simultaneously, and avoids overwriting existing data by default—making its practical value clear. OpenAI-compatible APIs and editable prompts support multiple models and self-hosted deployments in a lightweight but extensible implementation. With 0 GitHub Stars currently, community feedback and evidence of reliability are still limited, so it is best suited to early adopters willing to test and tune it themselves. Before deploying to production, validate permissions, rate limits, and failure retries, and refine prompts to match your site's tone. Adding caching/queues, logging and cost tracking, support for more SEO plugins, and unit tests would help with larger-scale adoption and community growth.