WordPress AI News and Updates - CourtneyR.dev's 2nd Brain

WordPress AI News and Updates 🔗 **Source:** [WordPress AI News and Updates](https://www.youtube.com/watch?v=sa827ct7W_I) > [!note]+ Document Notes > tags: [wordpress-ai, plugin-development, core-development] > author: Jonathan Bossenger > source: YouTube > reference: [https://www.youtube.com/watch?v=sa827ct7W_I] > date created: 2024-12-19 10:30 > > # 🏷️ Title: WordPress AI News and Updates >  ## 🧰 Action Inbox - [H1] Install Plugin Check Plugin and test naming tool — evaluate current plugin names #plugin-development - [H2] Try Abilities API JavaScript client with Gutenberg plugin — test local implementation #core-development - [H3] Install AI Experiments plugin v1.1 — explore content generation features #ai-tools ## 🧭 Table of Contents - [📝 Summary](#-summary) - [📖 Main Content](#-main-content) - [🗒️ Notes](#️-notes) - [📚 References](#-references) - [🗣️ Paraphrase](#️-paraphrase) - [📰 Newsletter](#-newsletter) - [✅ Action Items](#-action-items) - [🧪 Hypotheses](#-hypotheses) - [📊 Methodology](#-methodology) - [❓ Questions](#-questions) - [🔮 Future Research](#-future-research) - [💡 Implications](#-implications) - [🧠 Learning Styles Assessment](#-learning-styles-assessment) - [📊 Accessibility Assessment](#-accessibility-assessment) - [🌍 DEIB Assessment](#-deib-assessment) - [✅ Logical Validity](#-logical-validity) - [🛡️ DARVO Analysis](#-darvo-analysis) - [🎯 Conclusion](#-conclusion) ### 🔗 References - [Plugin Check Plugin](https://wordpress.org/plugins/plugin-check/) - [WordPress AI Client](https://github.com/WordPress/wordpress-ai-client) ## 📝 Summary - WordPress AI ecosystem continues expanding with new tools for plugin authors, core development, and content generation - Plugin Check now includes AI-powered naming validation to reduce review conflicts and trademark issues - Abilities API JavaScript client available in Gutenberg plugin, likely shipping with WordPress 7.0 - Core AI team released AI Experiments plugin v1.1 showcasing practical AI integrations - WP Bench provides objective evaluation of language models for WordPress development tasks ## 📖 Main Content The WordPress AI landscape is rapidly evolving with several significant updates across core development and community tools. The Plugin Check plugin now incorporates AI-powered naming validation, helping authors avoid common submission issues before review. The Abilities API, introduced in WordPress 6.9, gains a JavaScript client through the Gutenberg plugin, enabling direct browser-based execution without REST API calls. The Core AI team's experimental plugin demonstrates practical content generation features, while WP Bench offers objective model evaluation for WordPress-specific development tasks. ## 🗒️ Notes #### ⏱️ [00:00](https://youtu.be/sa827ct7W_I?t=0) — [H1] Plugin Check Namer Tool Introduction - Key idea: AI-powered plugin naming validation prevents review conflicts - Evidence: "You can now use a new tool called plugin check namer to help plugin authors evaluate plugin names for potential conflicts, trademark issues, and naming best practices" #### ⏱️ [02:15](https://youtu.be/sa827ct7W_I?t=135) — [H2] Configuring AI Provider for Name Checking - Key idea: Tool requires AI provider configuration with API keys - Evidence: "You select the provider you want to use, you pop your API key in, and then you select the model that you want to use" #### ⏱️ [03:20](https://youtu.be/sa827ct7W_I?t=200) — [H3] Testing Plugin Name "WPA SDK Demo" - Key idea: Tool correctly identifies WordPress naming guideline violations - Evidence: "The display name WPA SDK demo violates naming guidelines because it uses WP as a standard information for WordPress which is discouraged" #### ⏱️ [04:10](https://youtu.be/sa827ct7W_I?t=250) — [H2] Successful Name Validation Example - Key idea: Personal branding in plugin names passes validation checks - Evidence: "No issue detected. You can call it Jonathan's AI SDK demo. That's fine because I'm Jonathan" #### ⏱️ [05:00](https://youtu.be/sa827ct7W_I?t=300) — [H1] Abilities API JavaScript Client Update - Key idea: WordPress 6.9 Abilities API gains JavaScript client through Gutenberg plugin - Evidence: "With the JavaScript client, you don't have to do any of that. you just use the core functionality that the JavaScript API makes available" #### ⏱️ [06:45](https://youtu.be/sa827ct7W_I?t=405) — [H2] Testing JavaScript Abilities API - Key idea: JavaScript client enables direct ability execution from browser console - Evidence: "Get abilities work. So now if I go let ability... Call get site info. Call get environment info" #### ⏱️ [08:30](https://youtu.be/sa827ct7W_I?t=510) — [H3] Command Palette Integration - Key idea: Abilities integrate with WordPress command palette for workflow execution - Evidence: "If I open up the command pallet and I go command J for example, now I can run abilities and workflows" #### ⏱️ [09:15](https://youtu.be/sa827ct7W_I?t=555) — [H2] Core Post Management Abilities - Key idea: New pull request adds comprehensive post management through abilities - Evidence: "Add abilities for create post, get post, find posts and update posts" #### ⏱️ [10:30](https://youtu.be/sa827ct7W_I?t=630) — [H1] AI Experiments Plugin v1.1 Release - Key idea: Core AI team showcases practical AI building blocks for developers - Evidence: "The goal of this plug-in is just to showcase the different building blocks within WordPress and how they work" #### ⏱️ [12:00](https://youtu.be/sa827ct7W_I?t=720) — [H2] Content Generation Features Demo - Key idea: Plugin demonstrates excerpt and title generation from existing content - Evidence: "Generate an excerpt... Discover how the WordPress AI client simplify adding AI features to your plugins and themes" #### ⏱️ [14:20](https://youtu.be/sa827ct7W_I?t=860) — [H1] WP Bench Model Evaluation Tool - Key idea: Open source tool evaluates language model WordPress development competency - Evidence: "How well do language models actually understand WordPress... evaluates how well AI models understand WordPress development from core AIs, coding standards to plug-in architecture" #### ⏱️ [16:45](https://youtu.be/sa827ct7W_I?t=1005) — [H2] WP Bench Results and Model Comparison - Key idea: Benchmark shows Claude Opus currently leads WordPress development understanding - Evidence: "It shows a knowledge percentage. It shows a correctness percentage... Currently Claude Opus is the better of the models" ## 📚 References - [Plugin Check Plugin](https://wordpress.org/plugins/plugin-check/) - [WordPress Abilities API](https://make.wordpress.org/core/tag/abilities-api/) - [AI Experiments Plugin](https://github.com/WordPress/ai-experiments) - [WP Bench Repository](https://github.com/WordPress/wp-bench) ## 🗣️ Paraphrase - Plugin authors now have AI-powered name validation to avoid common review issues and trademark conflicts before submission - The Abilities API introduced in WordPress 6.9 gains JavaScript client functionality through Gutenberg plugin, eliminating REST API dependencies - WordPress Core AI team demonstrates practical AI integration patterns through experimental plugin showcasing content generation capabilities ## 📰 Newsletter WordPress developers gain powerful new AI tools this week! The Plugin Check plugin now includes intelligent naming validation to catch trademark issues before review, while the Abilities API gets JavaScript client support through Gutenberg. The Core AI team's experimental plugin showcases real content generation features, and WP Bench helps you choose the best AI model for WordPress development. Key benefits: Faster plugin approval process, streamlined core API usage, and objective AI model evaluation for WordPress-specific tasks. Resources: [Learn WordPress](https://learn.wordpress.org/), [Developer Docs](https://developer.wordpress.org/), [WP-CLI](https://wp-cli.org/) ## ✅ Action Items #### WordPress - Core: evaluate JavaScript abilities client for 7.0 roadmap - Plugins: integrate AI naming validation into review workflow - Core AI: expand experimental plugin feature set based on community feedback - Documentation: create abilities API JavaScript client guides - Community: promote WP Bench for model evaluation standards #### Community - Contributors: test plugin naming tool with existing plugins - Developers: experiment with abilities JavaScript API patterns - Maintainers: provide feedback on AI experiments plugin implementations ## 🧪 Hypotheses - AI-powered plugin naming validation will reduce review team workload and approval times - JavaScript abilities client will increase API adoption among front-end developers - Standardized AI model benchmarking will improve WordPress development tool selection - Practical AI demonstration plugins will accelerate community adoption patterns ### 📊 Methodology - Track plugin review rejection rates before/after naming tool deployment - Monitor abilities API usage patterns comparing REST vs JavaScript clients - Survey developer model preferences against WP Bench scoring results ### ❓ Questions - How will JavaScript abilities client affect WordPress performance and security models? - What additional AI building blocks should the experimental plugin prioritize next? - How frequently should WP Bench evaluations run to track model improvements? ### 🔮 Future Research - Investigate automated plugin code quality assessment using AI - Explore real-time model performance tracking for WordPress-specific tasks - Study community adoption patterns for AI-enhanced development workflows ### 💡 Implications - Practice: Developers gain direct browser-based access to WordPress core functionality - Policy: AI naming validation establishes precedent for automated review processes - Education: Experimental plugins provide concrete learning examples for AI integration - Funding: Model evaluation tools enable data-driven decisions for AI service investments ### 🧠 Learning Styles Assessment Reading level estimate: 10th Grade #### 👀 Visual learners - Present: Screen recordings of tool interfaces, console outputs, test results - Improve: Architecture diagrams showing abilities API flow, comparison charts #### 🔊 Aural learners - Present: Detailed walkthrough narration, live testing commentary - Improve: Separate audio summary highlighting key technical concepts #### 🗣️ Verbal learners - Present: Clear technical definitions, step-by-step explanations - Improve: More analogies comparing AI tools to familiar WordPress concepts #### 🧩 Kinesthetic learners - Present: Hands-on testing demonstrations, interactive tool usage - Improve: Written setup checklists, reproducible configuration steps #### 🧠 Logical learners - Present: Sequential tool testing, cause-effect relationships shown - Improve: Performance metrics, decision trees for model selection #### 🧑‍🤝‍🧑 Social learners - Present: Community tool context, team collaboration mentions - Improve: Discussion prompts for testing results, peer review suggestions #### 🧘‍♀️ Solitary learners - Present: Individual testing approach, personal experimentation - Improve: Self-assessment questions, independent learning paths #### 🌿 Naturalistic learners - Present: Evolution of WordPress AI ecosystem connections - Improve: Tool lifecycle thinking, sustainable development practices ### 📊 Accessibility Assessment #### Screen reader semantics - Present: Clear heading structure, descriptive content flow - Improve: Alt text for any visual demonstrations, ARIA labels for code examples #### Keyboard navigation - Present: Standard video controls, logical content order - Improve: Transcript navigation, code snippet accessibility #### Contrast and readability - Present: High contrast video interface, clear narration - Improve: Transcript formatting, code syntax highlighting review #### Text resizing and reflow - Present: Video responsive playback - Improve: Transcript reflow testing, mobile-friendly code examples #### Media and captions - Present: Audio narration available - Improve: Full video captions, downloadable transcript #### Forms and interactions - Present: Clear demonstration of form interactions - Improve: Accessible form validation examples in demos ### 🌍 DEIB Assessment #### Cultural breadth - Present: Global WordPress community context, open source focus - Improve: Multi-language AI model considerations, international naming guidelines #### Inclusive language - Present: Technical terminology well-defined, respectful tone - Improve: Plain language alternatives, avoid technical assumptions #### Neurodiversity support - Present: Structured content, consistent pacing - Improve: Content warnings for rapid screen changes, clearer section breaks #### Representation - Present: Focus on technical merit, community-driven development - Improve: Diverse use case examples, varied developer perspectives #### Psychological safety - Present: Encouraging experimentation tone, learning-focused approach - Improve: Explicit "safe to try" messaging, error handling guidance ### ✅ Logical Validity #### Strong reasoning - Deductive strength: Technical demonstrations support claims about tool functionality - Inductive strength: Model comparison results support general WordPress AI conclusions - Causal reasoning: Clear cause-effect relationships between tools and developer benefits - Analogical strength: WordPress ecosystem parallels help explain AI integration patterns #### Fallacies - Appeal to authority: Minor over-reliance on "core team" endorsement without technical analysis - Hasty generalization: Limited testing sample for drawing broad tool effectiveness conclusions #### Tone of voice - Overall tone: Enthusiastic but technically grounded, educational focus with practical examples ### 🛡️ DARVO Analysis - Deny: No denial patterns present - Attack: No attacks on individuals or competing approaches - Reverse victim/offender: No victim/offender dynamics - Emotional cues and rhetoric: Positive enthusiasm for technical progress, collaborative community spirit ## 🎯 Conclusion WordPress AI tooling reaches new maturity with practical solutions for common developer challenges. Plugin naming validation, JavaScript abilities client, and model benchmarking tools provide concrete value while experimental plugins demonstrate integration patterns. The focus on developer experience and objective evaluation suggests sustainable AI adoption rather than hype-driven implementation. ## 🧱 Refactor Appendix ### 📚 Literature Split Pad Purpose: Stage CEE units that become Literature notes. Split level: H3 ### [H1] Plugin naming validation reduces review friction > [!quote] WordPress AI News and Updates by Jonathan Bossenger via YouTube. Published 2024-12-19. > "Check your plug-in names for potential conflicts, trademark issues, and naming best practices" Addresses common plugin review bottlenecks through AI validation. [view highlight](https://youtu.be/sa827ct7W_I?t=135#H1) #plugin-development #ai-validation ID:rw_plugin_naming ### [H2] JavaScript abilities client eliminates REST dependencies > [!quote] WordPress AI News and Updates > "You just use the core functionality that the JavaScript API makes available" Simplifies WordPress core API integration for frontend developers. [view highlight](https://youtu.be/sa827ct7W_I?t=300#H2) #abilities-api #javascript ID:rw_js_abilities --- ### 🔖 Atomic Split Pad (Zettel Candidates) Purpose: Promote reusable claims into Atomic notes. Split level: H4 #### [Z1] AI model benchmarking enables objective tool selection - WordPress-specific evaluation criteria matter for development tasks - WP Bench provides standardized comparison methodology - Links: [[AI-Development-Tools]] [[WordPress-Core-API]] #### [Z2] Experimental plugins accelerate community AI adoption - Concrete implementation examples reduce integration barriers - Core team demonstrations establish best practice patterns - Links: [[Developer-Experience]] [[Community-Building]] --- ### ✏️ Fleeting Queue Purpose: Capture quick actions for Fleeting notes. - [F1] Test plugin naming tool with current projects — validate before submission #fleeting #plugin-check - [F2] Install Gutenberg plugin for abilities API testing — compare with REST approach #fleeting #abilities-api - [F3] Benchmark current AI models using WP Bench — update development toolchain #fleeting #model-evaluation --- ## Full Transcript [00:00] So today [00:04] I want to have a look at a bunch of AI [00:07] related news and updates around the [00:09] WordPress space. The very first one is [00:13] in the plug-in check plugin. So if you [00:15] don't know what the plug-in check [00:16] plug-in is, it's a plugin you can [00:18] install on your WordPress site [00:22] and use it to check your plugins for [00:25] submission to the plug-in directory. [00:27] So the big change, the big AI related [00:29] change that happened is that you can now [00:32] use a new tool called plug-in check [00:34] namer to help plug-in authors evaluate [00:37] plug-in names for potential conflicts, [00:38] trademark issues, and naming best [00:40] practices. So naming your plug-in is [00:42] something that comes up a lot in plug-in [00:45] reviews [00:49] and basically they check for things like [00:50] whether you're using trademarks or [00:53] whether you are using common terms. Um, [00:56] and so now you can use this plug-in [00:58] check name tool. You can put in the name [01:01] of the plug-in you want to evaluate. You [01:04] can enter your author or brand name if [01:05] you own the trademark. And you can check [01:08] if your proposed plug-in name has a [01:12] higher chance of being accepted by the [01:14] plug-in review team. So you select the [01:17] provider you want to use, you pop your [01:19] API key in, and then you select the [01:21] model that you want to use. You save [01:23] your changes and then you can use that [01:24] model and that provider to check your [01:27] plug-in names. But I'm just going to do [01:28] the one today and we'll see what [01:30] happens. So now if I go to the plug-in [01:33] check namer. I've actually got a couple [01:35] of plugins that I'm busy working on for [01:36] demo purposes. So I'm going to use one [01:39] of those. So it's possible naming [01:41] issues. The display name WPA SDK demo [01:44] violates naming guidelines because it [01:45] uses WP as a standard information for [01:48] WordPress which is discouraged in the [01:49] plug-in repository. And so there is my [01:52] analysis complete. So I can then go, [01:53] okay, this is not going to work. So what [01:55] about if I say something like Jonathan's [01:59] AI SDK demo? Let's see what that does. [02:02] Okay, so this one it seems to have been [02:04] happy with the results. But there it [02:05] says cool. No issue detected. You can [02:07] call it Jonathan's AI SDK demo. That's [02:09] fine because I'm Jonathan. So that's [02:12] obviously why it doesn't have a problem [02:13] with it. And there you go. So now I know [02:15] that this is a plug-in name that I can [02:16] use. So now I can go and rename my [02:17] plugin accordingly. Now, it doesn't mean [02:21] that your plug-in name will be accepted, [02:24] but it just means there's a higher [02:25] chance that it would be accepted, less [02:28] of a chance of somebody complaining [02:29] about it. Okay, let's move on. The other [02:31] thing that I discovered, [02:34] if you're following WordPress core [02:36] development in WordPress 6.9, [02:39] the abilities API was introduced to [02:41] WordPress. One thing that we didn't get [02:43] right for 6.9 was that the abilities API [02:46] has a JavaScript component. And what [02:49] that essentially means is you can use [02:51] the abilities package in your JavaScript [02:53] code and then you can execute abilities [02:56] from JavaScript. At the moment, if [02:58] you're just using core and you're using [03:00] abilities in core, you have to basically [03:02] query one of the rest API endpoints [03:04] using API fetch or one of those things. [03:08] With the JavaScript client, you don't [03:09] have to do any of that. you just use the [03:11] core functionality that the JavaScript [03:13] API makes available. So I discovered [03:18] this was merged in Gutenberg version [03:22] 22.0 as far as I know. So that means if [03:27] you have a WordPress install, a [03:29] WordPress 6.9 install and you don't have [03:32] any plugins active, but you've installed [03:35] the Gutenberg plugin, you should be able [03:37] to use the JavaScript API. So I'm going [03:40] to try that today and see if that does [03:42] indeed work. So to do that, what I'm [03:45] going to do is I'm going to install the [03:48] Reutenberg plugin. [03:53] And then once this has been installed, [03:58] going to see if I can do what I want to [03:59] do. [04:05] See if that works. [04:07] Okay, [04:12] pi. [04:15] Okay, so that seems to have worked and [04:17] there's executability. So I should have [04:23] get [04:25] abilities. Get abilities. Yes, get [04:27] abilities work. So now if I go let [04:30] ability [04:32] I can't spell no abilities equals [04:38] should be get abilities [04:43] and [04:46] console log abilities. [04:50] So there we go. Call get site info. Call [04:53] get environment info. Pretty awesome. [04:54] There are my abilities. So now I should [04:57] be able to go let result equals [05:01] execute ability and let's pass in [05:05] core get site info. [05:11] That should give me a result. I'm [05:12] supposed to console log [05:16] result. I should get a bunch of [05:17] information. [05:19] There's the result. And there we go. [05:21] There's the admin email and all that [05:22] stuff from my local install. So that's [05:24] cool. That means two things. That means [05:27] you can start using Abil's JavaScript [05:29] API client now if you have Gutenberg [05:30] plugin installed and it means this will [05:33] probably get shipped with WordPress 7.0. [05:34] And the other thing it means is that if [05:36] this pull request that we were looking [05:37] at earlier, it's adding experimental [05:39] abilities and workflow pallets to the [05:42] command pallet. If I open up the command [05:44] pallet and I go command J for example, [05:47] now I can run abilities and workflows. [05:49] So it picks up what abilities are [05:51] available and then executes them and [05:54] then does this. [05:56] So that's cool. I'm glad to see that [05:58] happening. So while we're there, there [06:01] is a pull request open to add more [06:02] abilities to WordPress core. So [06:05] basically core post management abilities [06:07] and the idea is to add abilities for [06:10] create post, get post, find posts and [06:12] update posts. There's a big discussion [06:14] around how this should work, but [06:17] essentially you'll be able with the [06:18] JavaScript client, you'll be able to run [06:20] this from browser. So you can run it [06:22] from anywhere, editor, any plugins, any [06:25] themes, whatever the case may be. [06:31] And the whole point is to eventually [06:33] make sure that everything that WordPress [06:34] does is available. If you are interested [06:36] in this development, I highly recommend [06:38] checking it out because I think it's [06:40] cool. I think the ability to be able to [06:42] trigger any of this functionality from [06:43] AI is cool. So, this one is more of a [06:46] check this out update. The WordPress [06:48] core AI team has released version 1.1 of [06:53] the AI experiments plug-in. And the goal [06:56] of this plug-in is just to showcase [07:00] the different building blocks within [07:02] WordPress and how they work. So, at the [07:04] moment, some of the ideas is we want to [07:06] have an AI playground. They want they [07:08] want to build a content assistant. They [07:11] want to build a site agent. They want to [07:13] show some workflow automation. And so [07:15] the goal is just to create a plug-in [07:17] that showcases what is possible with all [07:18] of these tools so that users can install [07:21] it, can check it out, can see what it [07:22] does, developers can play with it, can [07:24] look at the code, maybe use the code for [07:26] some some of their own workflows. So I'm [07:29] going to install this now and just see [07:30] what it can do. So this thing is telling [07:33] me the battery is dead, but I have it [07:35] plugged into power. [07:37] I'm not sure why it thinks the battery [07:38] is dead. Okay, you're stuck with this [07:40] view now. I don't know why that [07:42] happened. It is very annoying. [07:45] A generate exit button in the exit [07:46] panel. So, [07:50] generate an exit. Oh, there. Okay. All [07:52] right. So, let us do this. Okay. [07:57] Let's take the title out there [08:01] and the title in there. [08:04] So, let's say I'm working in a post. [08:05] Let's save the draft. [08:08] And now I can generate an excerpt. [08:12] Okay, so it's doing some things. [08:15] And there we go. Discover how the [08:16] WordPress AI client simplify adding AI [08:18] features to your plugins and themes by [08:20] centralizing a API API integrations and [08:24] credential. Oh, it's actually got more. [08:26] Okay. And credential management. It's a [08:28] whole excerpt. That's pretty cool. [08:30] There's a regenerate excerpt option. I [08:33] like it. It looks like title might be [08:35] possible. [08:37] Generates title from content. Okay. So, [08:39] let's see. Oh, there there's a generate [08:41] button. Okay. So, if we remove the [08:43] title, [08:46] then we can go generate [08:49] and it takes the content obviously uses [08:51] the AI client, passes it off to one of [08:53] the models, and then returns some [08:55] titles. So, that's cool. I'm going to [08:59] select that one for example. And there [09:00] it is. I think it's pretty cool to see [09:02] these things in place. So you can [09:03] generate a title from content. You can [09:05] generate an excerpt. It looks like it's [09:07] going to support image generation. Okay, [09:10] cool. So the rest is probably still [09:11] coming, but that's pretty cool. Glad to [09:13] see that work is progressing. Kudos to [09:15] the team working on it. I look forward [09:17] to seeing some more of these things [09:19] becoming available. [09:22] The last thing I wanted to do is this [09:24] thing WP Bench. So my understanding [09:26] about WPBench is it answers the [09:29] question, how well do language models [09:31] actually understand WordPress. [09:35] So this is important because if you're [09:36] using a coding tool or you're using an [09:39] AI agent, you're going to want to know [09:40] which is the best one to use for [09:42] WordPress development. Okay? [09:45] And what WPBench does, it evaluates how [09:49] well AI models understand WordPress [09:50] development from core AIs, coding [09:52] standards to plug-in architecture and [09:54] security best practices. It is available [09:56] as an open source repository. It's part [09:58] of the core AI team. It uses Python as [10:01] far as I can see, and I don't know if my [10:03] Python configuration is set up properly [10:06] on my machine. [10:08] I also need to create ENV files with [10:10] provider keys. [10:12] So, I'm going to have to take some [10:13] things off screen to do that. But [10:15] apparently, this is how it all works. [10:18] So, I'm guessing you clone it [10:22] and then you can run it. So, let's do [10:25] that. [10:27] Okay. So, is run from the root. [10:30] I completely missed that. [10:34] Okay. Dpbench. [10:38] And let's go. [10:41] All right. Oh, here we go. Okay. So, it [10:43] was so it starts development site starts [10:46] running. [10:48] Okay. [10:56] Oh, wow. It's actually checking like all [10:58] the models probably that I configured in [10:59] the YAML file. Yeah. Okay. So, it's four [11:01] and four mini claims and opus. So, [11:04] obviously you can configure other [11:05] models. You probably could also [11:08] configure like [11:10] open source models and run them locally [11:12] or from hugging face. Okay, so that is [11:16] done. So it shows a knowledge [11:17] percentage. [11:19] It shows a correctness percentage. [11:21] Currently doesn't show anything about [11:23] quality and then an overall percentage. [11:25] And currently Claude Opus is the better [11:29] of the models, which is what I've heard [11:30] from folks who are using these things. [11:33] But that's a pretty cool thing to do [11:34] like once a week or once a month or [11:36] whatever and just see and make sure [11:37] you're using the right the right [11:39] information. So that's WP Bench. It's a [11:42] great way to [11:46] determine [11:48] whether you're using the right model. [11:50] Okay, with that folks, I'm going to call [11:51] it a day. I have marked off all the [11:53] things I wanted to mark off. So, I hope [11:56] you have a wonderful rest of your [11:57] Saturday, whatever time it is, and I [12:00] will see you all going well next week, [12:02] Friday, as usual, at the normal time or [12:05] on YouTube or on the internet. All [12:07] right, have a good one. Bye.