What is MCP?

Model Context Protocol (MCP) is an open standard developed by Anthropic that creates a universal bridge between AI assistants and external tools. Instead of switching between your DAM and an AI chat, MCP lets your AI assistant search, organize, and manage your Razuna assets directly — securely and in real time.

With the Razuna MCP server your AI assistant can:

• Search and browse files across all your workspaces
• Upload, move, delete, and organize digital assets
• Find visually similar images using AI-powered similarity search
• Create and manage notes, documents, and personal AI memories
• Manage tags, collections, and custom metadata
• Generate share links and analyze file content with AI
• Track file usage analytics — downloads, views, and trends
• Use Collect+ as your personal clipboard for saving items from anywhere

How It Works

The MCP server acts as a secure bridge between your AI assistant and your Razuna account. Here's the flow:

1. Authentication — You provide an access token so the AI acts as you, seeing only the workspaces and files you have permission to access.
2. Natural language — You describe what you need in plain English (e.g., "Find all product photos tagged 'hero-image' in the Marketing workspace").
3. AI processing — The assistant translates your request into the right MCP tool calls.
4. Secure execution — The MCP server calls the Razuna API on your behalf, enforcing your workspace permissions and tenant isolation.
5. Results — Data flows back to the AI, which formats a helpful response for you.

Search & Find Assets

Locate any file across all your workspaces using natural language — no more hunting through folders.

Example prompts

• "Search for all product photos in the Marketing workspace"
• "Find files tagged 'brand-guidelines' uploaded this month"
• "Show me all PDF documents in the Legal folder"
• "Search for images related to our summer campaign"

The AI uses search_files behind the scenes, with support for keyword matching, tag filtering, file type filtering, and workspace scoping. Results include file metadata, thumbnails, and download URLs.

Visual Similarity Search

Find images that look similar to a reference image — powered by AI visual embeddings.

Example prompts

• "Find images similar to file [ID]"
• "Show me photos that look like our hero banner image"
• "Find visually similar product shots to this reference"

The AI uses find_similar_images which leverages AI-powered visual embeddings to find images with similar content, colors, and composition — even if they have completely different file names or tags.

Organize Files & Folders

Move, tag, and structure your digital assets through conversation.

Example prompts

• "Move all files tagged 'archive' to the Archive folder"
• "Create a new folder called 'Q3 Campaign' in the Marketing workspace"
• "Add the tag 'approved' to files [ID1], [ID2], and [ID3]"
• "Show me the folder tree for the Design workspace"
• "Delete all files in the Temp folder that are older than 30 days"

The AI combines create_folder, move_file, bulk_move_files, add_tag_to_files, and get_folder_tree to keep your asset library organized.

Upload & Ingest Assets

Bring new assets into Razuna and save web content — right from your AI conversation.

Example prompts

• "Upload this file to the Product Photos folder"
• "Save the URL https://example.com/image.png to my Design workspace"
• "Upload a file and tag it with 'hero-image' and 'campaign-2025'"

The AI uses upload_file for direct uploads and save_url to ingest content from the web. Files are automatically indexed and available for search immediately.

AI Memory & Knowledge

Store and recall personal knowledge, decisions, and context that persists across sessions — your AI never forgets important details.

Example prompts

• "Remember that the brand color is #FF6B35 and our font is Inter"
• "What do you know about our image naming conventions?"
• "Store a memory: approved vendors for stock photography are Unsplash and Shutterstock"
• "Recall everything you know about our Q3 campaign assets"

The AI uses store_memory and recall_memory to maintain a personal knowledge base that survives across chat sessions. Memories are scoped to your user and searchable by meaning (semantic search), not just keywords.

Notes & Documents

Create structured notes and documents stored alongside your assets — perfect for project briefs, style guides, and meeting notes.

Example prompts

• "Create a note called 'Brand Guidelines Summary' in the Marketing workspace"
• "List all notes tagged 'brief' in the Campaign workspace"
• "Search my documents for anything about social media image dimensions"
• "Update the 'Asset Naming Convention' document with the new rules"

The AI uses create_note, search_notes, create_document, and search_documents. Notes are team-visible workspace content; documents support rich text with tagging and full-text search.

Tags & Metadata Management

Organize your taxonomy and enrich assets with custom metadata — at scale.

Example prompts

• "List all tags in the Marketing workspace"
• "Create a tag called 'campaign-2025' with description 'Assets for 2025 campaign'"
• "Copy all tags from the Design workspace to the Marketing workspace"
• "Set the custom field 'license_type' to 'royalty-free' for file [ID]"
• "Show me the custom field schema for this workspace"

The AI uses tag tools (list_tags, create_tag, copy_tags_to_workspace) and custom field tools (get_custom_fields, get_file_custom_fields, set_file_custom_fields) to manage your taxonomy and metadata.

Sharing & Collections

Create share links and organize assets into collections for external collaborators.

Example prompts

• "Create a share link for the Q3 Campaign folder"
• "Create a collection called 'Client Review — March Deliverables'"
• "Generate a public share link for file [ID]"

The AI uses create_share_link and create_collection to produce shareable URLs and curated asset groups for clients and stakeholders.

Collect+ Personal Clipboard

Collect+ is your personal clipboard — save items from anywhere and process them later within Razuna.

Example prompts

• "Show me everything in my Collect+ inbox"
• "Search Collect+ for items about logo designs"
• "Move this Collect+ item to the Brand Assets workspace"
• "Save this URL to Collect+ for later review"
• "Create a note from this Collect+ item"

The AI uses list_collectplus_items, search_collectplus, move_collectplus_to_workspace, and save_url_collectplus to manage your personal clipboard. Items can be promoted to full workspace assets when ready.

Usage Analytics & Insights

Understand how your assets are being used — downloads, views, geographic reach, and device breakdowns.

Example prompts

• "Show me the usage stats for file [ID]"
• "How many times was our logo downloaded this month?"
• "What are the top viewed assets in the Marketing workspace?"

The AI uses get_file_usage_stats which returns download counts, view counts, unique users, usage trends over time, geographic distribution, and browser/OS analytics.

Multi-Step Workflows

Chain multiple actions together in a single conversation to handle complex tasks.

Example prompts

• "Search for all images tagged 'draft', move them to the Review folder, and add the tag 'pending-approval'"
• "Find all files uploaded this week, create a note summarizing them, and generate a share link for the folder"
• "Analyze file [ID] with AI, update its description with the analysis, and tag it appropriately"
• "List all workspaces, then show me the folder tree for each one"

The AI orchestrates multiple tool calls in sequence — searching, analyzing, tagging, moving, and sharing — to complete multi-step tasks that would normally require many clicks.

• Claude Desktop — Anthropic's desktop app (macOS, Windows, Linux)
• Claude.ai — Claude web app (Pro, Max, Team, Enterprise plans via custom connectors)
• ChatGPT — OpenAI's web interface (Pro/Plus with Developer Mode)
• Cursor — AI-powered code editor with built-in MCP support
• VS Code with Copilot — GitHub Copilot's MCP integration
• Raycast — Productivity launcher with MCP extensions
• Windsurf — AI IDE with native MCP support
• Claude Code — Anthropic's terminal-based coding agent
• Codex CLI — OpenAI's local coding agent

Step 1: Get Your Access Token

Your access token authenticates the AI as your user — it only sees workspaces and files you have permission to access.

1. Log in to Razuna
2. Go to Profile → Access Tokens
3. Click Create New Token
4. Give it a name (e.g., "Claude Desktop") and copy the token

Important: Copy the token immediately — you won't be able to see it again after creation.

Step 2: Connect Your AI Client

Choose your AI client below and follow the instructions.

Claude Desktop

Edit your Claude Desktop configuration file:

• macOS: ~/Library/Application Support/Claude/claude_desktop_config.json
• Windows: %APPDATA%\Claude\claude_desktop_config.json
• Linux: ~/.config/Claude/claude_desktop_config.json

Add the following configuration:

{
  "mcpServers": {
    "razuna": {
      "url": "https://mcp.razuna.com/sse",
      "headers": {
        "Authorization": "Bearer YOUR_ACCESS_TOKEN"
      }
    }
  }
}

Replace YOUR_ACCESS_TOKEN with the token you created in Step 1. Restart Claude Desktop completely (Cmd+Q / Alt+F4, then reopen).

Claude.ai (Web)

Custom connectors via remote MCP are available on Pro, Max, Team, and Enterprise plans:

1. Open Settings in the sidebar
2. Navigate to Connectors and select Add custom connector
3. Set Name to Razuna
4. Set URL to https://mcp.razuna.com
5. Authenticate with your access token when prompted

Cursor

Add to your project's .cursor/mcp.json file:

{
  "mcpServers": {
    "razuna": {
      "url": "https://mcp.razuna.com/sse",
      "headers": {
        "Authorization": "Bearer YOUR_ACCESS_TOKEN"
      }
    }
  }
}

Enable the Razuna MCP server in Cursor Settings. You should see a green dot when connected.

VS Code with GitHub Copilot

Add to your .vscode/mcp.json file:

{
  "servers": {
    "razuna": {
      "url": "https://mcp.razuna.com/sse",
      "headers": {
        "Authorization": "Bearer YOUR_ACCESS_TOKEN"
      }
    }
  }
}

Or use the Command Palette: MCP: Add Server → HTTP, enter the URL and name.

ChatGPT

1. Enable Developer Mode: Settings → Connectors → Advanced settings → Developer mode
2. In the Connectors tab, click Create
3. Set Name to Razuna and URL to https://mcp.razuna.com
4. Set Authentication to Custom Header, header name Authorization, value Bearer YOUR_ACCESS_TOKEN

Raycast

1. Run the Install Server command
2. Set Name to Razuna, Transport to HTTP
3. Set URL to https://mcp.razuna.com/sse
4. Add header Authorization: Bearer YOUR_ACCESS_TOKEN

Windsurf

Add to your mcp_config.json:

{
  "mcpServers": {
    "razuna": {
      "serverUrl": "https://mcp.razuna.com/sse",
      "headers": {
        "Authorization": "Bearer YOUR_ACCESS_TOKEN"
      }
    }
  }
}

Files (15 tools)

Full lifecycle management of digital assets — upload, search, move, delete, tag, transform, and more.

Tool	Description
upload_file	Upload a file to a workspace folder.
get_file	Get detailed metadata for a single file including name, size, type, tags, and preview URLs.
update_file_metadata	Update a file's name, description, or keywords.
delete_file	Move a file to trash.
move_file	Move a single file to a different folder.
bulk_move_files	Move multiple files to a folder in a single operation.
bulk_trash_files	Trash multiple files at once.
bulk_restore_files	Restore multiple files from trash.
add_tag_to_files	Add a label/tag to one or more files.
remove_tag_from_file	Remove a label/tag from a file.
save_url	Save a URL as a file in Razuna.
transform_image	Transform an image — resize, crop, rotate, convert format, and adjust quality.
move_files_to_workspace	Move files to a different workspace.
list_folder_files	List files in a folder with pagination.
empty_trash	Permanently delete all files in the trash.

Folders (3 tools)

Browse and manage your folder structure.

Tool	Description
get_folder_tree	Get the complete folder hierarchy for a workspace.
get_folder	Get details of a specific folder.
create_folder	Create a new folder in a workspace.

Workspaces (2 tools)

Browse and inspect your Razuna workspaces.

Tool	Description
list_workspaces	List all workspaces accessible to you.
get_workspace	Get full details of a specific workspace.

Search (4 tools)

Powerful search across files, knowledge base, and visual content.

Tool	Description
search_files	Primary file search with keyword matching, tag filtering, file type filtering, and workspace scoping.
find_similar_images	AI-powered visual similarity search — find images with similar content, colors, and composition.
search_knowledge	Combined search across both notes and memories in a single query. Recommended for "find everything about X" queries.
knowledge_graph_search	Alias for search_knowledge — searches across all knowledge stores.

Tags (6 tools)

Manage your tag taxonomy across workspaces.

Tool	Description
list_tags	List all tags in a workspace.
get_tag	Get details of a specific tag.
create_tag	Create a new tag with name, description, and color.
update_tag	Update a tag's properties.
delete_tag	Delete a tag.
copy_tags_to_workspace	Copy tags from one workspace to another — useful for standardizing taxonomy.

Memory (8 tools)

Personal AI knowledge store — decisions, context, and learnings that persist across sessions.

Tool	Description
store_memory	Save a new memory with title, content, and optional tags.
recall_memory	Semantic search across your memories — finds results by meaning, not just keywords.
search_memory	Alias for recall_memory.
read_memory	Read a specific memory by ID.
update_memory	Update a memory's title, content, or tags.
delete_memory	Delete a memory.
list_memories	List all memories with optional tag filtering and pagination.
suggest_memory_tags	Get tag suggestions to reuse existing tags and avoid duplicates.

Notes (7 tools)

Structured workspace notes — project briefs, meeting notes, and team documentation.

Tool	Description
create_note	Create a note in a workspace folder.
read_note	Read a note's full content.
update_note	Update a note's title, content, or tags.
delete_note	Delete a note.
list_notes	List notes with optional folder and tag filtering.
search_notes	Semantic search across all notes in a workspace.
suggest_note_tags	Get tag suggestions for notes.

Documents (7 tools)

Rich-text documents stored in workspace folders — style guides, asset specs, and reference material.

Tool	Description
create_document	Create a new document with rich-text content.
read_document	Read a document's full content.
update_document	Update a document's title, content, or tags.
delete_document	Delete a document.
list_documents	List documents with optional filtering.
search_documents	Full-text search across all documents.
suggest_document_tags	Get tag suggestions for documents.

Advanced (6 tools)

Collections, share links, custom metadata fields, and AI content analysis.

Tool	Description
create_collection	Create a curated asset collection.
create_share_link	Generate a shareable link for a file or folder.
get_custom_fields	Get the custom metadata field schema for a workspace.
get_file_custom_fields	Get custom field values for a specific file.
set_file_custom_fields	Set or update custom field values on a file.
analyze_file_content	AI-powered content analysis — get descriptions, detected objects, text extraction, and suggested tags.

Collect+ (8 tools)

Personal clipboard — save, search, and manage items collected from anywhere.

Tool	Description
list_collectplus_items	List items in your Collect+ inbox.
search_collectplus	Search Collect+ items by keyword.
read_collectplus_item	Read the full content of a Collect+ item.
update_collectplus_item	Update a Collect+ item's metadata.
delete_collectplus_item	Delete a Collect+ item.
move_collectplus_to_workspace	Promote a Collect+ item to a full workspace asset.
create_note_from_collectplus	Create a workspace note from a Collect+ item.
save_url_collectplus	Save a URL to your Collect+ inbox.

Usage (1 tool)

Asset analytics and usage tracking.

Tool	Description
get_file_usage_stats	Get comprehensive usage statistics — downloads, views, unique users, trends over time, geographic distribution, and browser/OS analytics.

Authentication Methods

The Razuna MCP server supports multiple authentication methods to work with any MCP client:

Access Token (Recommended)

Pass your access token in any of these ways:

• Authorization: Bearer YOUR_TOKEN header (standard, recommended)
• x-access-token: YOUR_TOKEN header
• RAZUNA_ACCESS_TOKEN environment variable (for stdio/local mode)

SSE Metadata

When connecting via SSE transport, tokens can also be passed in the connection metadata:

• accessToken or access_token in the SSE metadata payload
• authorization in the SSE metadata payload

Security

All requests are scoped to your user account. You only see workspaces and files you have permission to access. Tenant isolation is enforced server-side — there is no way to access another account's data. The MCP server is available exclusively to paid Razuna accounts.

Be specific

The more specific your prompt, the better the result. Instead of "show me files," try "show me all PNG images in the Marketing workspace tagged 'hero-image'."

Use context chaining

Within a single AI conversation, the assistant remembers previous results. You can say "now move those files to the Archive folder" without repeating file IDs.

Leverage AI memory

Store important conventions, brand guidelines, and naming rules as memories. The AI will recall them automatically in future sessions, keeping your work consistent.

Use combined search

When you're not sure where information lives, use search_knowledge to search across both notes and memories in a single query.

Tag suggestions before creating

Before creating new tags, ask for tag suggestions. This helps reuse existing tags and keeps your taxonomy clean.

Bulk operations for efficiency

For large-scale changes, use bulk tools like bulk_move_files and bulk_trash_files rather than processing files one at a time.

No tools appear / AI can't connect

• Verify your access token is correct and hasn't expired
• Ensure the MCP server URL is exactly https://mcp.razuna.com/sse
• Restart your AI client completely after configuration changes
• Check that your Razuna account is on a paid plan

"Access denied" on specific workspaces

The MCP server enforces your user permissions. If you can't access a workspace via MCP, verify that you're a member of that workspace in your Razuna account settings.

Slow responses for large workspaces

Workspaces with thousands of files may take a moment to search. Use specific filters (tags, file type, folder) to narrow results. Use list_folder_files with pagination rather than searching the entire workspace when you know which folder to look in.

Memory not recalling correctly

Memories use semantic search — phrase your recall queries naturally. If you stored "brand color is #FF6B35", searching for "what is our brand color" will find it. Use tags on memories for more precise filtering.