Pipeline Publisher: Automated AI Content Creation Plugin for WordPress

Stop writing from scratch and start orchestrating. Pipeline Publisher transforms your ideas into a steady pipeline of automated, intelligently generated content right inside your WordPress dashboard. Install the lightweight plugin today, connect your preferred AI models seamlessly, and watch your blog grow.

Download Free on WordPress.org

Powerful AI Prompt Chaining and Spintax Prompt Variables

Create complex, automated content pipelines without the layout clutter of traditional autobloggers. Pipeline Publisher lets you easily chain multiple prompts together by carrying your generated article text forward into subsequent steps. Using the dynamic :OUTPUT: placeholder, your core article automatically feeds the next stages of execution to generate highly relevant post titles, matching featured images, and clean blog excerpts.

To keep your content fresh and unpredictable, you can use native spintax variables directly inside your prompt fields. Simply list your variations using a standard format like [ hats | shirts | shorts ], and the plugin will randomly select an option every time the pipeline executes. This allows you to chain diverse topics and writing tones together for completely organic, hands-free content generation.

Hands-Free Content on Your Schedule

Keep your website active and updated completely on autopilot using the built-in scheduling dashboard. Pipeline Publisher gives you the flexibility to execute a specific prompt at precise intervals or configure the engine to pull randomly from a list of prompts you select. This ensures your blog maintains a diverse publishing footprint without requiring daily manual oversight.

Because the system is engineered specifically for the WordPress ecosystem, all automated tasks are handled by the native background scheduler (WP-Cron). This architectural choice guarantees reliable, worry-free execution that runs seamlessly in the background without slowing down your server or degrading your visitors’ page-load speeds.

Granular Operational Control & Smart Fallback Models

Take total command of your content generation workflows and API expenses. Pipeline Publisher places you firmly in the driver’s seat by offering developer-first settings, from custom system-level prompts to independent temperature and max token configurations for every task. You can dictate the exact rules, constraints, and creative boundaries for how your content is processed.

The plugin also features an intelligent model failover utility to keep your operating costs low. By specifying a comma-separated list of text or image models, the plugin always defaults to your preferred low-cost option first. If that model hits a rate limit or encounters an API failure, the engine automatically falls back to your secondary models in order, keeping your publishing pipelines moving without a hitch.

Setup, FAQ and Troubleshooting

Our helpful FAQ answers common setup, configuration and troubleshooting questions you may have while using Pipeline Publisher.

Can I use this plugin for free?

Yes. Because we route connections through your own WordPress Connectors, you can take advantage of the generous ongoing free tiers offered by providers like Google Gemini for text generation. If you decide to add advanced tasks like AI image generation, you can easily scale by adding API credits directly to your provider accounts on your own terms.

How do I configure my AI API key to work with Pipeline Publisher?

Pipeline Publisher doesn’t ask for your API keys directly; instead, it safely utilizes the Connectors API introduced in WordPress 7.0.

To configure your AI connections in WordPress 7.0 and newer:

Navigate to your WordPress dashboard and go to Settings > Connectors.
Locate your preferred provider (such as Google or OpenAI) and click to enter your API credentials.
Once saved and verified natively by WordPress, Pipeline Publisher will automatically detect the connection.

For a detailed walkthrough of how this core screen operates, you can refer to the official WordPress.org Core documentation.

How does the fallback model system save me money and keep the content pipeline flowing?

AI requests can fail if a specific model is overloaded or hits a temporary rate limit. Pipeline Publisher lets you input a list of preferred models in order of preference. The plugin will always try to use your preferred model first. If that request encounters an error, it seamlessly drops down to your fallback models to finish the task, ensuring your automated schedules never skip a beat.

To save money, simply enter your models into the plugin settings in a comma-separated list from least expensive to most expensive. We’ll only use your more expensive models if a request to your cheaper model fails.

By default the plugin attempts to use cheaper models first, before calling more expensive AI models.

What user roles can create and manage these pipelines?

To keep your site’s operational data secure, full plugin settings and API error logs are strictly restricted to site Administrators. However, to make it easy for content teams to collaborate, user roles with Editor privileges can fully create, edit, and schedule prompts without having access to your underlying system connections.

Where can I find the AI model names for manually specifying my preferred models in the plugin’s settings?

Because Pipeline Publisher routes requests directly through your configured WordPress Connectors, you can use any text or image model officially supported by your provider’s API.

The exact text strings (IDs) for models can be found in the official developer documentation for your chosen provider:

Google Gemini:

Visit the Google AI for Developers “Available Models” page, located at:
https://ai.google.dev/gemini-api/docs/models

OpenAI / ChatGPT Models:

Visit the OpenAI Developer platform documentation under “Models”, located at:
https://platform.openai.com/docs/models

Common IDs include gpt-4o or gpt-4o-mini.

If you have deleted your models within Pipeline Publisher and need to re-create your models you can use the following comma-separated lists of models:

Text Generation Models:

gemini-2.5-flash, gemini-2.5-flash-lite, gemini-2.5-pro, gpt-4o-mini, claude-3-haiku, gpt-3.5-turbo, gemini-1.5-flash, gemini-1.5-flash-8b, claude-3-5-haiku, gpt-4-turbo, gpt-4o, claude-3-sonnet, claude-3-5-sonnet, gemini-1.5-pro, o1-mini, o1-preview, o3-mini, gpt-4, claude-3-opus

Image Generation Models:

gemini-3.1-flash-image, gemini-2.5-flash-image, dall-e-3, imagen-3.0-generate-002, dall-e-2

How do I create my first prompt?

To create your first prompt, go to Pipeline Publisher > Prompts and click the Add New Prompt button at the top of the screen.

From within the prompt generation screen, you can configure the different options of how your content will be generated. Based on the prompt settings you configure, an AI generated post will be created and added to the Posts section of your blog.

The Post Text Prompt is a required field and will be used for generating the body of your blog post. All other prompts are optional but can be used to set up a truly automatic blog publishing pipeline.

Within each additional prompt you can use the :OUTPUT: text snippet to reference the content of the auto-generated article. This makes it easy to generate a custom featured image, title or excerpt based on the content automatically generated by your chosen AI agent. You can also choose to let your AI agent automatically pick a category for your post, or you can ensure that your new post always is assigned to a category you choose.

Use the Temperature setting to determine how creative and abstract your AI agent can be. Lower values will yield a more structured and deterministic writing style, whereas higher values will yield a more creative writing style.

You can also use the Max Tokens field to limit the tokens used by your AI agent for the post content. If you find that posts generated by a prompt are getting cut off mid sentence, you’ll need to increase your max token limit.

Once you’re done creating your prompt, click the Publish button within WordPress to save your changes. Then click the Generate Post Now button to test your prompt and have AI generate a new post for you instantly!

How do I use variables in my prompt?

Using variables in your prompt is easy. In any of your prompts, you can use our spintax variables syntax to have Pipeline Publisher randomly choose one value from a list.

Spintax variable lists start with an opening bracket [ and end with a closing bracket ]. Each value should be separated by the pipe | character.

For instance, if you wanted to write a short story but have the story focus on a different random animal each time, you would have a prompt like this:

Write a short story about a [ dog | cat | squirrel | rabbit | fox ] who works with all the forest animals to save the forest.

You can use as many different groups of variables as you want within a single prompt, which allows you to create some really specific and creative AI output.

How do I schedule a prompt to run automatically?

You can schedule a prompt to run automatically under the Pipeline Publisher > Post Scheduler menu. Simply create a new scheduler post and configure the options to run your desired prompt at the interval you choose. You can also select one or more prompts from a list and have one chosen at random for scheduled generation.

My AI generation requests are failing. How can I fix this?

There are a few reasons AI generation requests may fail:

AI provider rate or service limits
AI provider paywalls
Web host limitations or AI generation timeout

Sometimes issues with your chosen AI provider may prevent full generation of your AI content. For instance, some AI providers may impose rate limits or otherwise de-prioritize your AI generation requests, especially if using a free tier. This might come back as a 503 Service Unavailable response from your AI provider. Other AI features, such as image generation, may require the purchase of AI credits or otherwise not work on a free AI generation plan.

For information about what is causing your requests to fail you can go to the Pipeline Publisher > API Failure Logs screen, which will show you information about any errors encountered when trying to run your prompts.

In the event that AI requests are failing entirely and you do not see them in the API Failure Logs section, or in the event you get a browser timeout screen when running a prompt manually, it’s possible your AI generation chain is taking longer than is allowed by your web host. When running several prompts in a row it’s possible for requests to take longer than 30 seconds, which is where web hosts start to limit how long a request can run.

If you suspect your requests are timing out, you can go to Pipeline Publisher > Settings and enable the Set Custom Timeout setting. Not all web hosts will allow you to override the default system timeouts, but if your host does allow this via the PHP set_time_limit() function you can set a custom timeout value here.

In the event that setting a custom timeout value does not work, try reducing the number of prompts run for a single request. For instance, turning off automatic title, excerpt and category settings within a prompt may allow your server to complete the rest of the AI request without timing out.

How can I report a bug, request a feature or get additional support?

To report a bug or security issue with the plugin, or to suggest a feature, please use our contact form to get in touch with us.

If you need help using the plugin, we’re happy to offer paid support services, including plugin installation, configuration and troubleshooting. You can send us a message or schedule a free consultation with us to learn more.