Linux 软件免费装

在这里免费安装WordPress以及其他软件

Banner图

Pipeline Publisher

开发者 yourrightwebsite
更新时间 2026年7月10日 00:27
PHP版本: 8.3 及以上
WordPress版本: 7.0
版权: Expat
版权网址: 版权信息

标签

ai autoblogger prompt auto-poster content-generation

下载

1.0.0

详情介绍:

Pipeline Publisher is a structured AI content generator designed for automated blogging pipelines. Instead of relying on a single prompt to build an entire page, the plugin guides your content execution through a four-step generation loop to create a structured WordPress post. A core feature of Pipeline Publisher is its ability to chain your prompts together. The text generated by your main article is carried forward to subsequent steps in the pipeline. You can inject this generated text into your next prompts using a simple :OUTPUT: text snippet, giving you precise control over how the AI utilizes the previous step's data. For example, once a new post is generated, you can feed the output text of that post into a new prompt that handles title generation, like so: Based on the following article, generate a SEO optimized title of not more than 150 characters: :OUTPUT:

安装:

Prerequisites Plugin Installation and Configuration
  1. Ensure you have at least one AI service connected to your website in the Settings > Connectors interface added in WordPress 7.0.
  2. Upload the plugin files to the /wp-content/plugins/pipeline-publisher directory, or install the plugin through the WordPress plugins screen directly.
  3. Activate the plugin through the 'Plugins' screen in WordPress.
  4. Optionally, configure the plugin settings and your preferred AI models under the Pipeline Publisher > Settings menu.

屏幕截图:

  • The scheduler interface lets you run your prompt automatically as frequently as every 15 minutes.
  • Configure your preferred AI models and add custom system-level prompts in Settings.

升级注意事项:

1.0.0
  • Initial release.

常见问题:

How do I configure the plugin settings?

You can configure the plugin settings under the Pipeline Publisher > Settings sidebar menu inside the WordPress Dashboard.

Where can I find the correct AI model names to paste into the 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:

Always ensure you copy the exact lowercase technical ID (e.g., gemini-2.5-flash) rather than the public marketing name. You can list multiple models separated by a comma to set your fallback order. If you leave the models field blank, WordPress will attempt to pick an available default model from your connected AI providers to complete your AI task. However, when no model is specified, only a single API request will be attempted. If that model is overloaded or unavailable, your request will fail. Adding multiple models to the Preferred Models text box will result in each model being tried one at a time in the order you specify until one succeeds at completing your task or until all models have been tried. This allows you to try a cheaper model by default, but automatically fail over to an alternative model if the first choice is overloaded or unavailable.

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 core infrastructure built into WordPress 7.0+. To configure your connections:

  1. Navigate to your WordPress dashboard and go to Settings > Connectors.
  2. Locate your preferred provider (such as Google or OpenAI) and click to enter your API credentials.
  3. 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: https://make.wordpress.org/core/2026/03/18/introducing-the-connectors-api-in-wordpress-7-0/

Do I need to buy AI credits to use this plugin?

Depending on the AI model you are using, you may need to purchase AI credits through an AI provider to use some or all of this plugin. Google's Gemini offers a generous ongoing free tier via their AI Studio website which can be used for text and lightweight generation tasks. For image generation via Gemini, you will need to purchase AI credits from Google. Additional AI providers like OpenAI and Anthropic may provide one-time free credit allowances to help you get started, with the ability to purchase more credits to unlock more AI generation or advanced features.

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, save it using the Publish button within WordPress. Then click the Generate Post Now button to test your prompt and have AI generate a new post for you instantly!

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.

What level of access do different users have to Pipeline Publisher?

WordPress Administrators have full access to Pipeline Publisher, including the ability to create and edit prompts and schedules, change plugin settings and view AI API success and failure logs. Editors have the ability to create and edit prompts and schedules, but cannot change plugin-wide settings or view AI API logs.

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

There are a few reasons AI generation requests may fail:

  1. AI provider rate or service limits
  2. AI provider paywalls
  3. 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.

Where can I get help using this plugin?

If you need help using this plugin, there are several ways you can get help:

  1. Visit the Pipeline Publisher Website for additional troubleshooting tips and documentation.
  2. Post a help request on the WordPress.org Support Forums section for this plugin.
  3. Contact us directly for paid support, or to report a potential bug or security issue.

更新日志:

1.0.0