Cosmic MCP. Manage content, media, and object types via AI.
Gemini Works with every AI agent you already use
…and any MCP-compatible client
Just plug in your AI agents and start using Vinkius.
Cosmic (Headless CMS & Content API) MCP Server connects your content infrastructure directly to your AI agent. Manage all your content, media, and object types—from drafting blog posts to updating metadata—using natural language.
It gives your AI agent direct write access to your entire content model.
What your AI agents can do
Batch objects
Performs bulk operations on up to 25 objects in a single request.
Create media
Uploads new media assets or creates records for existing media.
Create object
Creates a brand new content object with defined metadata.
Your AI agent can create new content objects or update existing ones by passing specific metadata and content fields.
The agent can upload new media assets, delete old files, or modify the metadata associated with media already in the CMS.
You can define new object types or modify the structure of existing object types programmatically.
The agent retrieves specific objects, lists all objects of a certain type, or pulls the full history of revisions for any piece of content.
Send multiple object operations in one batch request, speeding up large-scale content migrations or updates.
Ask AI about this MCP
Supported MCP Clients
Gemini Waiting for input…
019e5d0dbatch objects
Performs bulk operations on up to 25 objects in a single request.
019e5d0dcreate media
Uploads new media assets or creates records for existing media.
019e5d0dcreate object
Creates a brand new content object with defined metadata.
019e5d0dcreate object type
Defines and adds a new structure (type) for content objects.
019e5d0dcreate revision
Adds a version (revision) to an existing object, tracking its history.
019e5d0ddelete media
Removes media assets from the bucket.
019e5d0ddelete object
Deletes a specific content object entirely.
019e5d0ddelete object type
Removes a defined content object type structure.
019e5d0dget agent conversation history
Retrieves the message history for a specific agent conversation.
019e5d0dget object
Retrieves the complete data and metadata for a single content object using its unique ID.
019e5d0dget object type
Retrieves the structure and definition of a specific content object type by its slug.
019e5d0dget revision
Retrieves the full details and metadata for a specific content revision.
019e5d0dlist agent conversations
Lists all agent conversations that have taken place.
019e5d0dlist media
Lists all media assets currently stored in the Cosmic bucket.
019e5d0dlist object types
Lists every content object type defined in the bucket.
019e5d0dlist objects
Lists all content objects, allowing you to filter the results using a query parameter.
019e5d0dlist revisions
Retrieves a list of all available revisions for a specific content object.
019e5d0dsend agent message
Sends a direct message to a specified AI Agent.
019e5d0dupdate media
Updates the metadata associated with a specific media file.
019e5d0dupdate object
Modifies the metadata or content of an already existing object.
019e5d0dupdate object type
Changes the definition or structure of an existing content object type.
Choose How to Get Started
Build a custom MCP for your own tools, or connect a ready-made integration from our catalog.
Build Your Own
Turn any API into an MCP. Import a spec, define Agent Skills, or deploy with MCPFusion.
- Import from OpenAPI, Swagger, or YAML specs
- Create Agent Skills with progressive disclosure
- Deploy to edge with MCPFusion framework
- Built in DLP, auth, and compliance on every call
- Real time usage dashboard and cost metering
- Publish to catalog or keep private
Make Your AI Do More
Start with Cosmic (Headless CMS & Content API), then connect any of our 4,700+ other servers whenever your AI needs more. One click, no limits.
- Use this MCP plus 4,700+ others, all in one place
- Add new capabilities to your AI anytime you want
- Every connection is secured and compliant automatically
- Track usage and costs across all your servers
- Works with Claude, ChatGPT, Cursor, and more
- New servers added to the catalog every week
What you can do with this MCP connector
Cosmic (Headless CMS & Content API) MCP Server connects your content infrastructure straight to your AI agent. You can manage all your content, media, and object types using natural language, giving your agent direct write access to your whole content model.
Create and Modify Content Objects: Your agent can create new content objects using create_object, or update existing ones' metadata and content via update_object. You can send a batch of up to 25 object operations at once with batch_objects. You'll also find that your agent can define a whole new content structure with create_object_type, or change an existing one using update_object_type.
To clean up, your agent can delete specific objects with delete_object, or remove an entire object type structure using delete_object_type.
Manage Media Files: The agent can upload new media assets with create_media or create records for existing ones. You can modify the metadata of media files already stored with update_media, and if you need to wipe something, the agent can delete media assets using delete_media. You can check what media's sitting in the bucket by calling list_media.
Audit and Retrieve Content: You can pull the full details and metadata for a single content object using get_object, or check the structure definition of any object type with get_object_type. When you need to see what objects exist, your agent can list all of them with list_objects, or list all defined object types with list_object_types.
You can also pull a list of all media assets with list_media and get a list of all agent conversations with list_agent_conversations.
Version Control: Your agent tracks content history by adding a version (revision) to an object using create_revision, and it can retrieve the full details for a specific revision using get_revision. You can also get a list of all available revisions for any content object with list_revisions.
Execution and Communication: The agent can send a direct message to a specified AI Agent using send_agent_message, and it can retrieve the message history for a specific agent conversation with get_agent_conversation_history. You'll also be able to list every agent conversation that's taken place by running list_agent_conversations.
How Cosmic MCP Works
- 1 Subscribe to the server and provide your Cosmic Bucket Slug, Read Key, and Write Key.
- 2 Your AI client reads the available tools (e.g.,
get_object,create_media) and recognizes the necessary parameters (e.g., object ID, media URL, content fields). - 3 The agent executes the required function call, writing the data directly to your live Cosmic CMS instance.
The bottom line is, your AI agent treats your entire CMS as a set of callable functions, letting you manage content without writing API calls.
Who Is Cosmic MCP For?
Content teams, product developers, and site architects need this. If you're tired of jumping between a content editor, a database console, and a local dev environment just to push a single metadata change, this is for you. It lets you manage the entire content lifecycle from a single conversation.
Uses the AI agent to draft new content objects, organize media assets, or audit existing content entries across multiple content types.
Quickly queries content structures or updates metadata in staging environments without leaving the code editor or writing boilerplate API calls.
Manages the content model by creating new object types or ensuring that existing content objects adhere to a specific metadata schema.
What Changes When You Connect
- Speed up content updates. Use
batch_objectsto process up to 25 objects in one call, avoiding slow, sequential API calls for large-scale metadata changes. - Maintain a clean content model. Use
create_object_typeandget_object_typeto programmatically define and verify the structure of your content before publishing. - Keep a full audit trail.
list_revisionsandget_revisionlet you view and manage every historical change to an object, so you never lose track of what was published when. - Work with media and content together. You can use
create_mediato upload files and thenupdate_objectto link the new media ID to a specific content object in the same workflow. - Reduce manual lookups. Instead of writing complex queries, just ask the agent to
list_objectsand filter by criteria. It handles the query parameters for you. - Handle content deletion safely. Use
delete_objectordelete_mediato ensure that when you remove content, you are deliberately removing it and not just marking it as inactive.
Real-World Use Cases
Updating an entire product line's metadata
A product team needs to update the pricing and key details for 15 product pages. Instead of calling update_object 15 times, they ask the agent to run batch_objects. The agent executes the batch call, updating all 15 records atomically.
Drafting a blog post with new images
A content manager starts by asking the agent to create_object for a new draft post. Then, they upload the featured image using create_media. Finally, they tell the agent to update_object, linking the new media ID to the draft post, solving the entire workflow in three conversational steps.
Restoring a corrupted article version
An editor notices an article has incorrect metadata. They ask the agent to list_revisions for the object. They identify the correct version ID, and the agent uses get_revision to pull the data, then executes create_revision to restore the desired state.
Schema audit and cleanup
A developer wants to check if any object type has outdated fields. They use list_object_types to see all available structures, and then use get_object_type on a specific slug to validate the required metadata fields against the current content objects.
The Tradeoffs
Calling individual list tools sequentially
Trying to find all content requires calling list_object_types, then calling list_objects with specific filters, and then calling list_media separately. This is a brittle, multi-step process that requires complex client-side state management.
→
Instead, ask the agent to list_objects and pass your filter criteria in the query parameter. The agent handles the necessary filtering and consolidation across the object types for you.
Manually managing object relationships
A user creates an object, then manually copies the ID to the update_object call, and then repeats this process for every related media asset. This is tedious and error-prone.
→
Use the agent to orchestrate the flow: first, run create_media to upload the asset. Then, tell the agent to update_object and reference the newly created media ID. The agent handles the linking.
Assuming data existence
A developer tries to update_object without first checking if the object ID is valid, leading to an API failure and requiring manual debugging.
→
Always start by using get_object to confirm the object exists and retrieve its current state before attempting any modification.
When It Fits, When It Doesn't
Use this if you need to manage the entire content lifecycle—from defining the schema (create_object_type) to creating content (create_object), linking media (create_media -> update_object), and running bulk cleanup (batch_objects). It’s the full control panel for your CMS.
Don't use this if your only need is to list conversation history, or if you just need to send a message to an agent (use list_agent_conversations or send_agent_message for those specific tasks). You need the content tools, not the agent tools.
Independent Platform Disclaimer: Vinkius is an independent platform and is not affiliated with, endorsed by, sponsored by, verified by, or otherwise authorized by Cosmic. All third-party trademarks, logos, and brand names are the property of their respective owners. Their use on this website is strictly for informational purposes to identify service compatibility and interoperability.
VINKIUS INFRASTRUCTURE
Cloud Hosted
Managed infra
V8 Isolated
Sandboxed per request
Zero-Trust Proxy
No stored credentials
DLP Enforced
Policy on every call
GDPR Compliant
EU data residency
Token Compression
~60% cost reduction
Works with Claude, ChatGPT, Cursor, and more
The Model Context Protocol standardizes how applications expose capabilities to LLMs. Instead of operating in isolation, your AI gains direct access to external platforms, live data, and real-world actions through secure, standardized connections.
This server provides 21 capabilities that interface natively with Claude, ChatGPT, Cursor, and any MCP client. No middleware. No custom integration required.
Available Capabilities
Content changes shouldn't require logging into three different dashboards.
Right now, updating a single piece of content means logging into the CMS UI, finding the object, checking the media library for the latest image, manually updating the fields, and then hitting 'Save'. If you need to update 20 items, you're clicking through 20 separate pages and copy-pasting IDs.
With this MCP server, you just tell your agent: 'Update all 'product' objects with this new headline and link the featured image from this folder.' The agent executes the necessary calls—using tools like `list_objects`, `update_object`, and `create_media`—and pushes the changes instantly.
Cosmic (Headless CMS & Content API) MCP Server: Full Content Control
The whole pain point of content is managing the structure. You manually define the type in the UI, then you manually link the media. This is what breaks when teams scale.
This server gives your AI client full control: it manages the object types with `create_object_type`, handles media with `create_media`, and ties it all together with `update_object`. You get a single, programmatic source of truth for your content.
Common Questions About Cosmic MCP
How do I use the `list_objects` tool with filters? +
You pass the filter criteria as a URL-encoded JSON query parameter. The agent handles the specific formatting required by the underlying API, so you just tell the agent what you need (e.g., 'list all objects where status is draft').
Can I use `batch_objects` to update multiple fields? +
Yes. The batch_objects tool is designed for bulk operations. You provide a list of object IDs and the updates you want to apply to each, making it efficient for mass metadata changes.
What is the difference between `update_object` and `create_object`? +
create_object makes a brand new piece of content. update_object changes content that already exists. You use the latter when you need to modify an existing object's metadata or body.
How do I track who changed a content object? +
You use list_revisions first to see all historical versions of the object. Then, you can use get_revision to pull the specific data from a past point in time.
What credentials do I need to use the `create_object` tool? +
You need a Cosmic Read Key and a Write Key to create objects. The MCP server handles authentication using these keys, so you don't need to worry about API calls. Just provide the keys when setting up the server.
How does `list_objects` handle complex filtering or querying? +
You pass a URL-encoded JSON string to the query parameter. This allows you to filter objects by specific metadata, date ranges, or object types. The AI client needs to format this JSON correctly for the tool to work.
If I delete an object using `delete_object`, can I restore it? +
No, deleting an object is permanent. However, you can use get_revision or list_revisions to retrieve previous versions of the object before it was deleted.
What is the scope of `batch_objects`? Does it cover all object types? +
The batch_objects tool handles multiple operations (up to 25) on existing objects. It's designed for high-volume updates, but you must specify the object types and IDs for the batch operation.
Can I filter objects by a specific type like 'blog-posts'? +
Yes! Use the list_objects tool and provide a JSON query string like {"type":"blog-posts"} in the query parameter to filter your results.
How do I update the content of an existing object? +
You can use the update_object tool. Provide the Object ID and the new values for title, metadata, or status to modify the entry.
Is it possible to see who made changes to an object? +
Yes, you can use the list_revisions tool for a specific object to see its history, and get_revision to inspect the details of a specific version.
Use it with your favorite AI tools
Connect this server to Cursor, Claude, VS Code, and more.
More in this category
Alphamoon
Extract data from documents using AI-powered OCR and intelligent document processing for contracts, invoices, and forms.
Svix
Manage your webhook infrastructure via Svix — orchestrate applications, configure endpoints, and monitor message delivery directly from any AI agent.
Render Alternative
Automate your PaaS infrastructure via Render — list your services, deploy code, check logs, and scale resources directly from any AI agent.
You might also like
Haystack (deepset Cloud)
Build and manage AI-powered search and RAG pipelines via deepset Cloud — search documents, run pipelines, and manage workspaces.
Replicate Extended
Automate machine learning workflows via Replicate — run models, manage predictions, and search for AI assets directly from any AI agent.
Anaplan
Manage Anaplan workspaces, models, and data actions (imports, exports, processes) directly via AI.