mcp s

Cloudinary MCP Servers expose a set of conversational interfaces that let you upload, transform, analyze, and organize media assets through natural language. Each server implements a specific domain: Asset Management for uploading and organizing assets, Environment Config for managing cloud settings and presets, Structured Metadata for defining and querying metadata fields, Analysis for AI-powered content analysis and tagging, and MediaFlows for building low-code, AI-assisted workflows. You can interact with these servers via MCP-enabled LLM tools, issue commands to start jobs, apply transformations, set metadata, or trigger automated pipelines, all from your AI application or chat interface. Use the remote MCP endpoints when you want Cloudinary-hosted capabilities, or run local MCP servers to customize behavior and test integrations before going to production.

Prerequisites:

• Node.js and npm (for local MCP servers)
• Access to your Cloudinary account and credentials (CLOUDINARY_URL or individual keys)
• Internet access to install npm packages

Step 1: Install Node.js and npm if you don’t have them

• Download from https://nodejs.org/ or use your system package manager.

Step 2: Configure Cloudinary credentials

• Obtain your Cloudinary API key, API secret, and cloud name from the Cloudinary dashboard.
• You can export CLOUDINARY_URL as cloudinary://api_key:api_secret@cloud_name or set individual environment variables by server (see configuration examples).

Step 3: Install and run local MCP servers

• Option A: Use the provided npm-based local configuration (recommended for testing).
  • Create a configuration JSON (as shown in mcp_config) to map each MCP server to its local start command.
  • Example for one server: npx -y --package @cloudinary/asset-management-mcp -- mcp start
  • Repeat for other servers (environment-config, structured-metadata, analysis, mediaflows) as needed.
• Option B: Use the remote Cloudinary MCP servers if you prefer a hosted setup.

Step 4: Start the servers

• Ensure your environment variables (CLOUDINARY_URL or individual keys) are set.
• Run your configured commands (see mcp_config) to start each server.

Step 5: Connect your MCP client

• Point your MCP client to either the local server endpoints or the remote Cloudinary endpoints described in the README’s Remote MCP Servers Configuration section.

Step 6: Verify and test

• Use the Cloudinary MCP documentation and the provided examples to validate uploads, transformations, analyses, and metadata queries through your AI toolchain.

Tips and common issues:

• If using CLOUDINARY_URL, ensure it has the correct format cloudinary://api_key:api_secret@cloud_name. Incorrect formatting will cause authentication failures.
• When running multiple local MCP servers, keep distinct ports and environments to avoid conflicts.
• For local testing, consider using a single CLOUDINARY_URL with test credentials and separate env files per server to avoid cross-talk.
• If a server fails to start, check Node.js compatibility and ensure the required npm package @cloudinary/<service>-mcp is accessible in your npm registry.
• The mediaflows server often benefits from additional Cloudinary presets or automations; consult the MediaFlows MCP docs for sample workflows.