# Obsidian Publish MCP for AI Agents AI Agent Connect > Obsidian Publish MCP lets your AI agent read your notes. It indexes files and crawls your navigation trees so your agent can find and summarize specific markdown content from your published vault instantly. ## Overview - **Category:** productivity - **Price:** Free - **Endpoint:** https://edge.vinkius.com/vk_preview_2g0EdU9eeaIsXiKX4gDTqfxPUZ4owDXPPYajo4f3/ai-agent-connect - **Tags:** markdown, vault-indexing, static-site, knowledge-base, note-taking, web-publishing ## Description The Obsidian Publish MCP lets your AI agent read your notes. You have a massive vault of notes, but finding that one specific detail in a sea of markdown is a chore. This tool changes that by giving your AI client a direct line to your Obsidian Publish site. Instead of copy-pasting chunks of text into a chat, your agent can browse your entire folder structure, find the right note, and pull the exact content you need. It's like giving your agent a seat in your personal library. Whether you're looking for a specific SOP in a corporate wiki or a character's backstory in a novel, your agent can navigate the tree, check the frontmatter for tags, and grab the raw text in seconds. You can even use it to audit your site's status or see how your navigation is mapped out. By connecting through the Vinkius catalog, you turn your static notes into a live knowledge base that your agent actually understands. It makes your notes searchable and actionable without you having to manually curate what the AI sees every single time. You no longer have to worry about whether your agent has the right context for a complex question. It can see the same hierarchy you do, meaning it can jump from a high-level overview to a granular technical detail without missing a beat. It turns your static markdown files into a live reference you can actually use. ## Tools ### get_metadata Grab the internal creation hashes and frontmatter properties for a specific page. It lets your agent see tags and internal links. ## Prompt Examples **Prompt:** ``` List all the files currently publicly available in my vault. ``` **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 ### Browse folder structures Your agent can see the full hierarchy of your published notes. ### Pull raw note content The agent grabs the exact text from a specific note for analysis. ### Extract frontmatter and tags It pulls hidden metadata and tags to help the agent understand note categories. ### Audit site configuration Your agent can check the status and mapping of your live site. ### Map all published files It gives your agent a complete inventory of every published file in your vault. ## Use Cases ### Cross-referencing research notes A researcher needs to find a specific quote in 500 notes. The agent uses list_files to see the scope and get_file to grab the quote. ### Querying a technical digital garden A developer wants to find deployment steps in a 'DevOps' folder. The agent uses list_navigation to find the folder and get_file to read the guide. ### Auditing internal SOP updates A manager wants to know if the 'HR Policy' note is updated. The agent uses get_metadata to check the last edit date and tags. ### Summarizing long documentation A team needs to summarize a 50-page SOP. The agent uses list_navigation to find the SOP and get_file to read the whole thing at once. ## Benefits - Stop copy-pasting notes into your chat. Use get_file to pull the exact content your agent needs. - Navigate complex folders easily. list_navigation lets your agent see your vault's structure. - Keep your knowledge organized. Use get_metadata to let your agent understand tags and internal links. - Verify your site status quickly. site_info lets you check if your notes are live and accessible. - Map your entire library in seconds. list_files gives your agent a full inventory of your published content. ## How It Works The bottom line is your AI gets a direct path to your notes without manual copy-pasting. 1. Subscribe to the Obsidian Publish MCP via Vinkius. 2. Enter your Obsidian Publish Site ID and an optional access token. 3. Ask your AI client to find, read, or summarize notes from your vault. ## Frequently Asked Questions **Can the Obsidian Publish MCP read my private notes?** Yes, it can read your private notes as long as you provide the required access token during the setup process. **Does this Connector let my AI agent edit my notes?** No, this Connector is designed for reading and summarizing content. It does not have permissions to edit or delete your notes. **How does the Obsidian Publish MCP help with research?** It allows your agent to quickly cross-reference different notes across your vault, making it much faster to find connections between topics. **Can I use Obsidian Publish MCP for my company's SOPs?** Yes, it is a great way to turn internal standard operating procedures into a queryable knowledge base for your team. **Does the Obsidian Publish MCP support binary assets?** Yes, the get_file tool is capable of retrieving both textual content and binary assets from your published notes. **How do I connect my vault to the Obsidian Publish MCP?** Simply subscribe to the Connector and enter your Obsidian Publish Site ID in the configuration settings. **Can the AI accurately recreate the folder structure of my vault?** Yes. When the model invokes `list_navigation`, it downloads the systematic JSON map of the entire public vault, thereby allowing it to understand the relationships and physical structure of your folders perfectly. **Does `get_file` automatically render complex Markdown components?** No rendering is performed on our end. The `get_file` tool returns pure raw Markdown (including frontmatter and obsidian syntax). Fortunately, LLMs (like Claude) are exceptionally adept at reading and reasoning over raw Markdown text structures. **If my Published site is private and requires a password, does it still work?** Yes, it works gracefully. If you configured Obsidian Publish with restricted read access, you can manually inject a persistent token into the setup, giving your AI agent the clearance to read securely behind the authentication wall.