# Obsidian Publish MCP > Obsidian Publish MCP lets your AI agent read every note in your published markdown knowledge base. It indexes your entire site structure and pulls the raw text content of specific documents so you can summarize years of notes or cross-reference complex technical specs instantly. You get deep access to structured, private, or public writing. ## Overview - **Category:** productivity - **Price:** Free - **Tags:** markdown, vault-indexing, static-site, knowledge-base, note-taking, web-publishing ## Description Your personal vault—the thousands of scattered markdown files across Obsidian Publish—is suddenly connected to your AI client. Instead of opening dozens of tabs and manually copy-pasting sections, your agent now sees the whole picture. It can build a functional map of your entire knowledge base by crawling the navigation structure and listing all available file paths. Need to summarize one specific document? You just ask for it, and the raw text streams directly into your chat window. This capability means you don't have to navigate complex local folders; you talk about your notes like they're in a database. Connecting this MCP through Vinkius gives your agent access to deep file content, technical metadata, and even site configuration details—all so you can finally build that smart chatbot or research assistant based on your own work. ## Tools ### list_files Lists all the file paths that have been explicitly published in your Obsidian workspace. ### get_file Retrieves the full, raw markdown text and any associated binary files for a specific note. ### site_info Checks global configuration details and styling used across your entire published site. ### list_navigation Visualizes the structured, linked map of markdown notes within your vault. ### get_metadata Extracts technical properties from a note, including tags, creation dates, and internal link logic. ## Prompt Examples **Prompt:** ``` Check the vault and list all the files currently publicly available. ``` **Response:** ``` According to the navigation tree, you have 14 distinct markdown entries published. Key documents include 'Index', 'Machine Learning Setup', and 'Project Specs'. ``` **Prompt:** ``` Read the contents of 'System Requirements 2026.md'. ``` **Response:** ``` I've pulled the content of 'System Requirements 2026.md'. The document includes heavy focus on horizontal scaling algorithms and Python-centric deployments. Would you like a condensed executive summary? ``` **Prompt:** ``` Fetch the metadata and tags applied to my 'Inbox' note. ``` **Response:** ``` The 'Inbox' note has the following metadata: generated on May 5th, tagged with `#todo` and `#ideas`, and links heavily to `Projects.md` and `Archived Content.md`. ``` ## Capabilities ### Map the entire published vault structure The MCP generates a structural map of all notes, showing how they link together in a navigable tree. ### Retrieve full raw markdown content You get the complete text and any associated assets for a specified note or file path. ### Extract internal document properties The system pulls technical data about notes, including creation dates, tags, and frontmatter fields. ### Verify site deployment details You can check the global configuration of your Obsidian Publish endpoint to confirm connectivity status. ## Use Cases ### Need a summary of all characters in my novel. Instead of opening every character profile note and manually reading through them, you tell your agent to use `list_files` to find all 'character' notes, and then uses those paths with the raw content retrieval to build a cross-referenced timeline for you. ### The team needs an SOP chatbot. You connect the MCP and ask it to ingest your entire documentation set. The agent can then answer, 'What is step 3 of the deployment process?' by querying the correct file's raw content. ### Checking if a note was properly categorized. You use `get_metadata` on a specific note to confirm its status, tags, and internal links. This confirms whether it's ready for publishing or needs more context before being referenced. ### Pulling all code examples from my technical documentation. You ask the agent to analyze your 'developer guide' vault, which uses `list_files` to identify all `.md` files, and then executes `get_file` on each one to extract only the markdown fenced code blocks for review. ## Benefits - Cross-reference complex ideas across your entire vault. Instead of hunting through files, you ask the agent to pull data from multiple sources and summarize them for you. - Audit your notes instantly. Use the MCP's ability to extract internal properties and metadata so you know exactly what tags or dates are attached to a piece of information. - Build corporate knowledge assistants. Feed your team's standard operating procedures into the AI, letting it rapidly answer questions about processes without requiring manual search queries. - Access raw content reliably. With `get_file`, you get the clean markdown source for any note, perfect for feeding specific code snippets or technical passages to other tools. - Understand site architecture. Use `list_navigation` to visualize your notes' relationship map, ensuring that no major section of your work is accidentally orphaned from the main structure. ## How It Works The bottom line is that your AI client treats your entire private markdown vault like one searchable knowledge graph. 1. Subscribe to this MCP and input your specific Obsidian Publish Site ID (and an optional token if the vault is private). 2. Tell your AI agent what you want to know—for example, 'Summarize all notes related to Q3 marketing goals.' 3. The agent executes multiple background calls to read the structure, pull raw content, and synthesize a single answer for you. ## Frequently Asked Questions **How does Obsidian Publish MCP handle private vaults?** You must provide your Obsidian Publish Site ID and an optional access token for the agent to read notes in private or restricted corporate vaults. This keeps your data secure. **Can Obsidian Publish MCP index code snippets?** Yes. The `get_file` tool retrieves the full raw markdown content, which includes any markdown-fenced code blocks and technical code examples housed in your notes. **Is this only for published sites?** No, it connects to your established Obsidian Publish environment. The MCP indexes files that are already structured within the publishing framework. **What is the difference between list_files and list_navigation?** List files gives a simple directory of every available file path. List navigation provides a visual, linked map showing how those notes relate to each other structurally within your vault. **Does Obsidian Publish MCP only read markdown?** It reads the raw markdown content and can retrieve associated binary assets that are embedded or referenced in your published notes.