# ButterCMS MCP for AI Agents MCP > ButterCMS MCP connects any AI agent to your entire publishing instance, letting you extract rich text and structured data directly from your Headless CMS. Use it to search blog posts by keywords, map content collections, or analyze internal taxonomy without ever opening the browser console. ## Overview - **Category:** developer-tools - **Price:** Free - **Tags:** headless-cms, content-api, taxonomy, data-extraction, knowledge-base ## Description This MCP lets your AI agent talk directly to your ButterCMS backend. Instead of manually navigating through dashboards or writing complex API calls, you simply ask for what you need—whether it's finding all articles about a specific topic or mapping out how different content types connect. Your agent reads structured data and rich text assets across your entire publishing site. It can pull detailed information from blog posts by slug, dig into the internal structure of tags and categories, and even enumerate complex custom page models spanning multiple layers of your website. Everything you need to analyze your content warehouse is accessible through natural conversation. When paired with Vinkius, you connect once and get access to this CMS intelligence alongside thousands of other data sources. ## Tools ### search_blog_posts Find articles and properties that contain specific keywords or topics within your blog posts. ### search_collection_field Identify specific arrays of filtered collections to understand which data groups are active on your site. ### get_page_layout Retrieve the exact structural matching verifying explicit UI routing ### get_post_details Fetch the full details and logging information for any specific post slug you provide. ### list_butter_authors Get a list of all authors who have published content through the CMS. ### list_butter_categories Analyze and retrieve every single defined category used across your entire site structure. ### list_global_collections Enumerate all structured content rules attached globally, showing what types of items are tracked. ### list_custom_pages Inspect and list deep internal arrays containing raw data models for custom-built pages. ### list_blog_posts List all available blog posts, identifying the general content routing spaces within the CMS limits. ### list_butter_tags Retrieve a comprehensive list of all global tags applied to content items for taxonomy mapping. ## Prompt Examples **Prompt:** ``` List all the main content categories we have defined on our site. ``` **Response:** ``` **CMS Categories Overview** We currently track 4 major content pillars: * 💻 Development (120 posts) * 📈 Marketing (78 posts) * 🎨 Design/UX (35 posts) * 💡 AI & Research (62 posts) You can narrow this down further by asking about tags or specific authors. ``` **Prompt:** ``` Can you find all articles related to 'growth hacking'? ``` **Response:** ``` I found three matching content pieces that discuss growth hacking: * **Title:** How to Growth Hack Your Codebase * *Author:* Jane Doe * *Published:* 2024-11-15 * *Tags:* #DevOps, #Strategy * *Snippet:* ...focusing on rapid iteration and minimal viable product development. * **Title:** Year 2 Recap * *Author:* John Smith * *Published:* 2024-12-01 * *Tags:* #Marketing, #Strategy * *Snippet:* ...highlighting the early success of our rapid growth initiatives. Would you like me to pull up the full text for any of these? ``` **Prompt:** ``` What are the main content types we track, especially custom ones? ``` **Response:** ``` We manage several structured data sources. In addition to standard blog posts and pages, we explicitly track: * **Global Collections:** This tracks rules for core product documentation. * **Custom Pages:** These hold specialized models like 'Client Case Studies' or 'Technical Specs', which have unique JSON outputs. * **Authors:** We also maintain a roster of all published authors and their associated content. ``` ## Capabilities ### Scan Blog Content by Topic Your agent searches the blog posts using keywords or structures to gather specific knowledge from your site's published articles. ### Map CMS Taxonomies You can analyze all configured tags, authors, and categories across the entire system to understand how content is grouped and structured. ### List Content Collections The MCP enumerates all globally attached structured rules and custom page models, giving you a map of your site's data structure. ## Use Cases ### Auditing Content Gaps for SEO A marketing manager needs to know if their competitor's topic is covered. They ask the agent to 'search_blog_posts' using a niche keyword phrase, immediately identifying relevant articles and content gaps. ### Mapping Site Architecture for Devs A headless engineer needs to verify if a specific custom page template is pulling its data correctly. They use 'list_custom_pages' to inspect the raw JSON output, confirming component hits and nested structures. ### Analyzing Content Ownership for PR A PR team needs to compile a list of all recent articles written by a specific author. They use 'list_butter_authors' combined with 'get_post_details' to quickly build a resource page. ### Understanding Content Relationships A content strategist wants to know the core pillars of their site. They ask the agent to list all categories and tags using 'list_butter_categories' and 'list_butter_tags', creating an instant, usable taxonomy map. ## Benefits - Instead of manually checking multiple CMS tabs, you can run a deep search using 'search_blog_posts' to gather knowledge across your entire site in one prompt. - You get immediate clarity on content structure; use 'list_butter_categories' and 'list_butter_tags' together to map out how every article is categorized and tagged. - Bypass the need for complex coding. Use 'list_custom_pages' to inspect raw page models, helping engineers debug why certain custom data isn't rendering right. - Quickly audit your content base by running 'list_butter_authors' to see who wrote what, which is crucial when proposing backlinking efforts to improve SEO. - Understand the site map instantly. Running 'list_global_collections' shows you every structured rule attached to your content, giving you a full view of your data architecture. ## How It Works The bottom line is you bypass manual database queries and get immediate access to deep content insights right where your AI agent works. 1. Subscribe to this MCP and provide your specific ButterCMS API Token from your project security settings. 2. Your AI agent connects using that token, giving it direct read access to all structured content within your CMS instance. 3. You ask a question—like 'Show me all articles about X' or 'What are our core categories?'—and the MCP returns clean, usable data directly into your chat window. ## Frequently Asked Questions **How does the ButterCMS MCP help me audit my website's content structure?** The MCP gives you a single view of your entire CMS data. You can use it to list all categories, tags, and collections in one go, letting you map out how every piece of content relates without manual effort. **I need to debug custom pages; is the ButterCMS MCP right for this?** Yes. The MCP lets engineers inspect raw page models using 'list_custom_pages'. This returns explicit JSON objects, so you can pinpoint why a component isn't rendering nested data correctly. **What if I only want to search for articles by a keyword?** You simply ask your agent to perform a full-text search. The MCP runs the query and pulls back all matching blog posts, giving you immediate results based on keywords or structural properties. **Can I use this MCP for SEO analysis in my content strategy?** Absolutely. You can easily run searches to check if target keywords already exist on your site using 'search_blog_posts'. This helps you avoid duplicating content and plan backlinking strategies. **Does the ButterCMS MCP handle author details, too?** Yes. You can request a list of all published authors using the dedicated tool, which is helpful for compiling resource pages or checking authorship history across your site.