How is this different from connecting Cloudinary MCP directly?

Connected directly, an agent sees every tool the server exposes and holds the credential itself. Through TrustGate you choose which tools each agent gets, and every call is logged against the agent that made it.

What if the Cloudinary MCP API changes?

We track the Cloudinary MCP schema, so the tool catalogue stays current. Tools you have already allowed keep working, and anything new stays denied until you enable it.

Can I give different agents different Cloudinary tools?

Yes. Access is set per agent: one can get a handful of read tools, another the full set. You grant the whole MCP or single tools within it.

Where can I see what my agents did in Cloudinary?

Every call is logged against the agent that made it, with the tool it used. Activity can be sent on to your own monitoring.

Do I need to run the Cloudinary MCP server myself?

No. TrustGate connects to the Cloudinary MCP server and sits in front of it. Your agent only talks to your TrustGate URL, never to Cloudinary directly.

Connect Cloudinary MCP to your AI agents, securely.

Q: Is Cloudinary MCP secure?

The server itself exposes 29 tools, and some of them can write to your Cloudinary account. Behind TrustGate an agent only sees the tools you allow, requests and responses are inspected, and every call is logged. Anything outside the allowlist is denied before it reaches Cloudinary.

One secured endpoint for every Cloudinary call your AI agents make, scoped per user, inspected, and logged.

Get started for free

Get a demo

Try Cloudinary MCP now

What your agents can ask once Cloudinary MCP is connected.

Select the tools your AI agents can call

29 tools from the Cloudinary MCP schema.

Showing 29 tools

list-imagesreadRetrieves a list of image assets, filterable by tags, prefix or specific public IDs.
list-videosreadRetrieves a list of video assets, filterable by tags, prefix or specific public IDs.
list-filesreadRetrieves a list of raw assets, filterable by tags, prefix or specific public IDs.
get-asset-detailsreadReturns the details of a single resource specified by its asset ID.
upload-assetwriteUploads images, videos or raw files. It accepts local paths through file://, remote HTTP and HTTPS URLs, base64 data URIs up to around 60 MB, private S3 or Google Storage buckets, and FTP addresses. Cloudinary analyses and indexes format, size, resolution and prominent colours on upload.
asset-renamewriteUpdates an existing asset's public ID and optionally other metadata, which changes its delivery URL.
asset-updatewriteUpdates one or more attributes of a specified asset by its asset ID, including metadata and tags.
delete-assethigh riskDeletes an asset using its immutable asset ID.
list-tagsreadRetrieves a comprehensive list of the tags currently applied to assets in the account.
manage-asset-tagswriteAdds, removes or replaces tags on multiple assets at once, addressing them by public ID.
manage-asset-contextwriteAdds or clears contextual metadata on multiple assets at once, addressing them by public ID.
manage-asset-metadatawriteAssigns structured metadata field values to multiple assets at once.
create-asset-relationswriteRelates an asset to other assets by their asset IDs, regardless of public ID.
delete-asset-relationswriteUnrelates an asset from other assets, specified by their asset IDs.
search-foldersreadLists the folders matching a search expression.
create-folderwriteCreates a new empty folder at the specified path in the media library.
move-folderhigh riskRenames or moves an entire folder, along with every asset it contains, to a new location.
delete-folderhigh riskDeletes a folder and all assets within it.
get-tx-referencereadReturns Cloudinary's transformation rules from the official documentation. Cloudinary marks this mandatory before creating, modifying or displaying a transformation.
transform-assetwriteGenerates derived transformations for existing assets using the explicit API with eager transformations.
generate-archivewriteCreates a downloadable ZIP or TGZ archive containing a specified set of assets from the product environment.
delete-derived-assetshigh riskDeletes derived resources by their derived resource ID.
get-generation-taskreadReturns the status of a generation task.
generate-imagewriteGenerates an image from a text prompt using AI models, with the model selectable per call.
generate-image-from-imageswriteGenerates an image guided by one or more reference images, for restyling or on-brand variants.
search-assetsreadA query interface to filter and retrieve assets and their details by search expression.
visual-search-assetsreadFinds images in the asset library by visual similarity or content rather than by metadata.
download-asset-backupreadDownloads a backup copy of an asset from its stored revision history.
get-usage-detailsreadReturns a report on product environment usage, including storage, bandwidth and account statistics.

What Cloudinary MCP exposes

Cloudinary MCP groups its 29 tools into 7 categories, each with a suggested starting policy.

All 29 tools as JSON, ready to paste.

Category	Covers	Access	Suggested policy
Assets	Uploading images, videos and raw files, listing images, videos and raw assets with filters, full details for one asset by its immutable asset ID, renaming, updating attributes and deleting	Read / Write	upload-asset accepts file:// local paths, remote URLs, base64 data URIs, private S3 and Google Storage buckets and FTP, so it can read from the host the server runs on. delete-asset works on the immutable asset ID
Metadata & Tags	Listing the tags in use across the account, applying tag commands to multiple assets at once, setting or clearing contextual metadata in bulk, setting structured metadata field values, and relating or unrelating assets by asset ID	Read / Write	The three manage- tools address assets in bulk by public ID, so one call changes many records. Tags and context often drive delivery and downstream automation, not just organization
Folders	Creating an empty folder, searching folders by expression, renaming or moving a whole folder with everything in it, and deleting a folder	Read / Write	delete-folder removes the folder and every asset within it. move-folder relocates an entire tree, which breaks any path-based delivery URL pointing at the old location
Transformations & Delivery	Reading Cloudinary's transformation rules from the official docs, generating eager derived transformations for existing assets, deleting derived resources, and building a ZIP or TGZ archive of a set of assets	Read / Write	Cloudinary marks get-tx-reference mandatory before creating or modifying a transformation. generate-archive packages assets for download, which is an egress path worth logging
AI Generation	Generating an image from a text prompt with a selectable model, generating an image guided by one or more reference images for restyling or on-brand variants, and checking the status of a generation task	Read / Write	Both generation tools call AI models and bill per call. Generation is asynchronous, so the task status tool is the read-only half of the pair
Search	A query interface to filter and retrieve assets and their details by expression, and finding images by visual similarity or content rather than by metadata	Read	Read-only. visual-search-assets matches on image content, so it can surface assets the caller could not have found by name or tag
Backup & Account	Downloading a backup copy of an asset, and a report on product environment usage including storage, bandwidth and account statistics	Read	Read-only. download-asset-backup pulls a stored revision out of the environment, so treat it as an egress path rather than a lookup

Connect Cloudinary MCP to your AI agent

Four steps, under ten minutes.

01
Sign in to NeuralTrust
Open the NeuralTrust platform, sign in and go to Getting started, then Connect your tools.
02
Choose your agent
When asked which agent you use, pick Claude Code, Codex, Cursor, Windsurf, Copilot or your own code.
03
Connect Cloudinary
Find Cloudinary in the tool list and connect it.
04
Connect your agent
Follow the steps on screen to add the config to your agent, then send the test prompt.

Permissions down to the single tool

You do not allow Cloudinary MCP or block it. You allow the exact tools each agent can call, one by one.

Cloudinary MCP	29 TOOLS AVAILABLE
Code review agent	4 tools, read only
Issue triage bot	9 tools
Platform team assistant	all 29 tools

Tool by toolConnecting the server is not the unit of access. Each tool is granted on its own.
Default denyA tool you have not enabled is never exposed to the agent.
Approval on writesTools that change or delete things can require a human to confirm.

Is Cloudinary MCP secure?

One endpoint instead of raw credentials. Every call is checked against policy and logged.

AI agentsConnect through your MCP client configuration.	NeuralTrustApply controls based on your configuration.	Cloudinary MCP29 documented tools in @cloudinary/asset-management-mcp v0.11.3, a local Node stdio server under the MIT licence. There is no OAuth: the manifest takes a cloud name, an API key and an API secret, with both credentials declared sensitive.
Coding assistants	Connection & credentialsper user	Assets8 tools
Internal copilots	Tool access policiesdeny by default	Metadata & Tags6 tools
Workflow automations	Request & response inspectioninline	Folders4 tools
Compatible MCP clients	Activity monitoringalerts and logs	All other tools11 tools