Upload, search, transform and organise your Cloudinary media library from a chat window
Connects assistants to a Cloudinary media library so they can upload files, run search and visual search, apply transformations, manage folders and tags.
A Cloudinary account. The hosted endpoint authorises over OAuth; running the npm package locally needs Node.js 18+ and a cloud name, API key and API secret.
Compare MCP servers by category, or browse hosted MCP servers if you prefer an endpoint over a local process.
Point the assistant at a folder of freshly imported assets and have it normalise public IDs, apply a consistent tag scheme, and flag anything that failed to upload.
Ask for a crop-and-optimise variant, look at the result, iterate on the transformation string in conversation, then copy the final string into your delivery code.
Search for every asset tagged for a campaign, confirm the list, and generate a single ZIP archive to send to an agency or partner.
Use visual search to locate an asset by describing what is in the frame when nobody remembers the filename or tag.
Check the linked publisher documentation and your connected client's tool list. Names and availability can vary with the server version and account permissions.
| Tool | What it does |
|---|---|
| upload-asset | Upload an image, video or raw file from a local path or remote URL, setting public ID, folder and tags at ingest time. |
| search-assets | Run a structured expression search across the media library with sorting, aggregation and pagination. |
| visual-search-assets | Find assets by visual similarity or image content rather than by filename or tag. |
| transform-asset | Apply a Cloudinary transformation string to an existing asset via the explicit API, with optional CDN invalidation. |
| get-tx-reference | Return Cloudinary's full transformation reference so the model can compose valid transformation strings. |
| get-asset-details | Fetch the full metadata record for one asset by asset ID. |
| asset-update | Update an existing asset — tags, contextual metadata, moderation state and other mutable fields. |
| asset-rename | Change an asset's public ID, optionally overwriting an existing target. |
| delete-asset | Delete an asset by asset ID. |
| list-images | List image resources in the product environment with filtering and pagination. |
| list-videos | List video resources in the product environment. |
| list-files | List raw (non-image, non-video) files. |
| list-tags | List the tags currently in use across the library. |
| create-folder | Create a new folder in the asset hierarchy. |
| search-folders | Search the folder tree by expression. |
| generate-archive | Build a ZIP or TGZ archive from a set of assets matched by tag, prefix or public ID. |
| generate-image | Generate a new image from a text prompt and store it in the library. |
| generate-image-from-images | Generate a new image conditioned on one or more existing assets. |
| get-generation-task | Poll the status of an asynchronous generation task. |
| get-usage-details | Report account usage — credits, storage, bandwidth and transformation counts. |
claude mcp add cloudinary --transport http https://asset-management.mcp.cloudinary.com/mcpA Cloudinary account. The hosted endpoint authorises over OAuth; running the npm package locally needs Node.js 18+ and a cloud name, API key and API secret.
| Variable | Description | Required |
|---|---|---|
| CLOUDINARY_CLOUD_NAME | Cloud name of the Cloudinary product environment you want the server to act on. Local runs only.e.g. demo | Optional |
| CLOUDINARY_API_KEY | API key for the product environment. Local runs only; the hosted endpoint uses OAuth instead.e.g. 123456789012345 | Optional |
| CLOUDINARY_API_SECRET | API secret paired with the API key. Local runs only. | Optional |
| CLOUDINARY_URL | Single-string alternative that encodes cloud name, key and secret together.e.g. cloudinary://key:secret@cloud-name | Optional |
| CLOUDINARY_MCP_APPS | Controls the interactive MCP UI apps (asset gallery, asset details, upload view). Set to 'none' to disable them.e.g. asset-gallery,asset-details | Optional |
Read, search and edit files inside allowed directories with the Model Context Protocol reference Filesystem server.
Retrieve version-aware library documentation and code examples with Context7 MCP, using resolve-library-id and query-docs.
基于 Model Context Protocol (MCP) 的 B 站视频与动态评论获取工具
A comprehensive proxy that combines multiple MCP servers into a single MCP. It provides discovery and management of tools, prompts, resources, and
Official MCP server for dbt (data build tool) providing integration with dbt Core/Cloud CLI, project metadata discovery, model information, and
AI coding agents spend most of their input tokens on tool output: build logs, test runs, git status, the same file read again after every edit. All of it goes into the context window raw, and it is…