# Mem AI Knowledge MCP > Mem AI (Knowledge Workspace) connects your agent directly to the Mem.ai knowledge platform. Use this MCP to turn conversation into structured, retrievable knowledge. You can create new notes instantly using Markdown, run deep semantic searches across every document, and organize everything into thematic collections. ## Overview - **Category:** productivity - **Price:** Free - **Tags:** knowledge-base, semantic-search, markdown-notes, information-retrieval, personal-knowledge-management ## Description Your agent takes full control of your personal and team knowledge in Mem.ai. Instead of manually navigating a dashboard or searching keyword by keyword, you simply talk to your AI client about the information you need. The system handles the retrieval process, using semantic search to find notes based on meaning, not just matching words. You can create new ideas—or 'mems'—in Markdown format directly from the chat window, keeping your thought process logged instantly. Need to organize a project? Establish thematic groupings and attach all relevant documents automatically. This MCP makes sure every piece of information you collect is searchable and contextually mapped. It's the core knowledge layer that lets you work with data conversationally, making it available through Vinkius, right alongside thousands of other tools. ## Tools ### create_mem Generates a new note using Markdown formatting, immediately indexing the content for future searches. ### list_mems Retrieves basic identifiers and the raw body content for all notes across your entire workspace. ### get_mem Fetches the full context, metadata, and specific details of a note using its unique ID. ### update_mem Overwrites the content of an existing note with new text provided by your agent. ### delete_mem Permanently removes a specific note document from the entire knowledge base; no recovery is possible. ### search_mems Performs an advanced search across all indexed notes, finding information based on underlying meaning rather than keywords. ### mem_it Triggers a quick capture shortcut to log links or snippets into the workspace without manual navigation. ### list_collections Queries and lists all thematic collections (folders) currently defined in your knowledge base. ### create_collection Establishes a new, logical container or grouping for related notes within the system. ### get_collection Retrieves metadata and details about a specific thematic collection you want to inspect. ### add_mem_to_collection Attaches an existing note into one of your defined thematic collections, structurally organizing it. ### list_collection_mems Queries and lists all individual notes that have been assigned to a specific collection. ## Prompt Examples **Prompt:** ``` Search my mems for anything related to 'quarterly business review' ``` **Response:** ``` Searching semantically… I've identified 3 relevant mems: 1) 'Q1 Review Goals', 2) 'Draft slides for QBR', and 3) a note about 'Revenue projections' that matches the context. Would you like me to retrieve the full content of the QBR slides draft? ``` **Prompt:** ``` Create a new mem with today's standup notes in Markdown ``` **Response:** ``` Mem created successfully! I've logged your standup notes into your workspace. New Mem ID: 'mem-12345'. Would you like me to add this mem to your 'Engineering Standups' collection? ``` **Prompt:** ``` List all my thematic collections in Mem ``` **Response:** ``` I've retrieved your collections: 'Project Alpha', 'Meeting Archives', 'Personal Research', and 'Reading List'. Which one would you like to explore or add a new note to? ``` ## Capabilities ### Build New Knowledge Notes Instruct your agent to generate a new note using Markdown formatting and save it immediately into the workspace. ### Semantic Information Search Ask the agent complex questions, and it will search all indexed notes for information relevant to the meaning of your query. ### Retrieve Specific Context Details Give the agent a specific note ID and have it fetch the full text body and associated metadata for precise details. ### Manage Thematic Collections Ask to list or create defined thematic groupings, keeping related notes structurally organized. ### Capture Quick Thoughts Trigger rapid capture blocks for links or raw ideas without needing to open a separate manual note-taking page. ### Update Existing Records Tell the agent to change content in an existing note while making sure historical context remains intact. ## Use Cases ### Synthesizing competitive research A marketing director needs to compare competitor feature sets from 50 different articles. Instead of manually reading everything, they ask their agent to use `search_mems` for 'pricing model' and 'API integration'. The agent pulls the most relevant notes and summarizes the differences into a single document. ### Onboarding new team members A project manager needs to train a new hire on an old product line. They use their agent to `list_collections` for 'Product X Manuals' and then ask the agent to summarize all notes attached to that collection, ensuring no critical step is missed. ### Debugging complex code issues A developer encounters a rare bug. They use their agent to `get_mem` for specific error logs from last week and then ask the agent to compare those notes against current documentation, instantly finding the relevant fix or workaround. ### Archiving completed projects At the end of a client engagement, an account manager uses their agent to `create_collection` called 'Client Alpha Wrap-up'. They then run multiple commands to attach all related meeting notes and deliverables into that single collection. ## Benefits - Stop relying on keyword matches. Use `search_mems` to find information based purely on meaning, allowing you to retrieve context from notes that don't explicitly mention the topic. - Maintain project boundaries by using `create_collection` and then assigning related documents with `add_mem_to_collection`. This keeps your knowledge organized by theme, not just date. - Capture ideas instantly. The `mem_it` tool lets you log quick links or raw thoughts into the system without breaking flow, treating every thought as indexed data. - Keep project logs current with confidence. Instead of overwriting notes entirely, use `update_mem` to append new meeting decisions while preserving the historical context needed for audits. - Understand your knowledge footprint by running `list_mems`. This gives you a clear inventory of every note and collection across your team's work. ## How It Works The bottom line is that you get an integrated layer over Mem.ai's powerful search and organizational tools, accessible through any MCP-compatible client. 1. Subscribe to this MCP and provide your Mem.ai API Key. 2. Connect your preferred AI client (like Cursor or Claude) to the Vinkius Catalog. 3. Use natural language prompts in your agent chat to initiate knowledge actions, such as requesting a search or asking to create new notes. ## Frequently Asked Questions **How does Mem AI (Knowledge Workspace) MCP handle search?** It uses semantic similarity, meaning it searches based on the *meaning* of your query rather than just matching keywords. This lets you find relevant context even if the notes use different terminology. **Can I organize my documents into specific groups with Mem AI (Knowledge Workspace) MCP?** Yes, you can establish thematic groupings using `create_collection` and then attach related notes to them for structural organization. This keeps your knowledge segmented by project or topic. **If I use mem_it, does it just save a link?** No, the `mem_it` tool generates an automated block that captures links or raw thoughts and indexes them immediately into your workspace, making them searchable alongside your full notes. **Is there a way to update old notes using Mem AI (Knowledge Workspace) MCP?** Yes, the `update_mem` tool allows you to modify existing content. It’s crucial to use this carefully, as it replaces the text, so always confirm what needs appending. **How do I get a full history of notes in Mem AI (Knowledge Workspace) MCP?** You can run `list_mems` to retrieve identifiers and raw bodies for all notes, giving you a comprehensive overview of your entire knowledge base.