Capacities MCP. Manage your entire knowledge graph via chat.
Gemini Works with every AI agent you already use
…and any MCP-compatible client
Just plug in your AI agents and start using Vinkius.
Capacities MCP Server gives your AI agent full control over your personal knowledge graph. Use it to build structured objects, save weblinks, and append notes directly into your Capacities spaces via conversation.
It lets you manage your entire object-based knowledge base without switching context.
What your AI agents can do
Add tag
Adds a structural tag to an object, grouping related graph items using predefined relations.
Create object
Creates a new, typed object within a Capacities space that adheres to specific graph rules.
Get object
Retrieves the full data for a specific object ID by traversing its root graph properties.
List all personal spaces in your Capacities account and retrieve the full metadata for object types within any space.
Create new, structured graph objects that comply with the specific rules and parameters of a target Capacities space.
Send short thoughts, summaries, or Markdown text directly into your mapped daily note log.
Save a web URL as a Weblink object, automatically generating and attaching a dynamic preview to your space.
Execute keyword searches across a specific Capacities space to find exact nodes based on titles or keywords.
Locate an image or media payload and bind it directly to an existing record or object.
Ask AI about this MCP
Supported MCP Clients
Gemini Waiting for input…
019d7569add tag
Adds a structural tag to an object, grouping related graph items using predefined relations.
019d7569create object
Creates a new, typed object within a Capacities space that adheres to specific graph rules.
019d7569get object
Retrieves the full data for a specific object ID by traversing its root graph properties.
019d7569get space info
Gets detailed metadata about a Capacities space, including its object types and property definitions.
019d7569get structures
Retrieves all defined object type blueprints (structures) within a Capacities space.
019d7569list spaces
Lists all top-level personal spaces available in your Capacities account.
019d7569lookup
Searches for content across a specific Capacities space using keywords or titles to find exact nodes.
019d7569save media
Finds and attaches an image or media file directly to an existing record or object.
019d7569save to daily note
Appends Markdown text content to today's mapped daily note, linking it to the content block.
019d7569save weblink
Saves a web URL as a Weblink object, which automatically generates a dynamic preview.
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 Capacities, 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
Capacities MCP Server - Knowledge Graphs for AI
Your AI agent gets full control over your personal knowledge graph. You'll use it to build structured objects, save weblinks, and append notes right into your Capacities spaces, all through conversation. You can manage your entire object-based knowledge base without switching context.
Map your knowledge graph structure. You can use list_spaces to list all your personal spaces and use get_space_info to get the full metadata for object types in any space. You'll see what kinds of data structures are available using get_structures and you can get detailed metadata about a space using get_space_info.
Build typed data objects. You'll create new, structured graph objects using create_object that follow the specific rules of a target Capacities space. You can also retrieve all the details for an existing object using get_object by traversing its root graph properties.
Quickly capture daily notes. You'll send short thoughts, summaries, or Markdown text straight to today's mapped daily note using save_to_daily_note, linking it right to the content block.
Store web links with previews. You can save a web URL as a Weblink object using save_weblink, which automatically generates and attaches a dynamic preview. You'll also save images or media files by finding and attaching them directly to an existing record or object using save_media.
Search specific data records. You'll execute keyword searches across a specific Capacities space using lookup to find exact nodes based on titles or keywords. You can also search for content across a specific Capacities space using lookup.
How Capacities MCP Works
- 1 Subscribe to the Capacities MCP Server and provide your Personal Token.
- 2 Give your agent a natural language request (e.g., 'Save this article link').
- 3 The agent executes the necessary tool (like
save_weblink) and writes the result back into your Capacities space.
The bottom line is, your AI agent operates inside your personal knowledge graph, making changes and retrieving data directly, without you touching the UI.
Who Is Capacities MCP For?
This is for the knowledge worker who struggles to keep disparate information organized. If you're a researcher drowning in notes and links, or a developer who needs to document system architecture quickly, this server lets your agent treat your entire workspace like a single, addressable database.
Uses the agent to capture research links and draft quick notes into the daily log without context switching.
Commands the agent to document system architectures by injecting structured Markdown objects into specific project spaces.
Asks the agent to search the workspace and extract specific object properties to brief them on a topic before a meeting.
What Changes When You Connect
- Build structured knowledge: Use
create_objectto build new typed graph objects that comply with your space's exact rules. Stop treating knowledge as loose text; start modeling it. - Context-aware searching:
lookupruns rapid keyword searches targeting explicit object hierarchies, so you don't have to guess where information lives. It finds the exact node you need. - Centralized daily logging: Send quick thoughts or summaries using
save_to_daily_note. Your agent logs it permanently to your daily note without you opening a new tab. - Rich, self-contained links:
save_weblinksaves a URL and automatically generates a preview. The link isn't just a dead end; it's a fully contextualized object. - Full data visibility:
get_structureslets you pull all the object type definitions from a space. This is key for developers needing to see the exact metadata parameters at work. - Organization at scale: Use
add_tagto structurally group related items. You can tag objects across different spaces, maintaining a cohesive graph relationship.
Real-World Use Cases
Archiving a research article
A researcher finds a useful article. Instead of opening a browser tab, copying the link, and pasting it into a note, they prompt their agent: 'Save this URL to my Research space as a Weblink.' The agent uses save_weblink, which captures the link and automatically injects the dynamic preview, saving time and context.
Documenting a new system module
A developer finishes a module. They ask their agent to 'Create a new System Component object in the Architecture space.' The agent uses create_object, ensuring the new record adheres precisely to the predefined structure parameters, which prevents data loss.
Quick meeting prep and data extraction
A consultant needs to brief on a project. They ask the agent: 'Search the Project space for all 'Key Milestones' objects and list their status.' The agent runs lookup and get_object to pull specific properties, giving the consultant a clean brief before the meeting.
Logging code fixes and notes
A developer finishes a bugfix. They prompt: 'Append the Python code and a summary of the fix to my daily note.' The agent uses save_to_daily_note, logging the code permanently and mapping the bugfix securely into their daily log.
The Tradeoffs
Using tags for everything
The user just throws everything into tags, making the graph messy and unstructured. They manually write: 'Tag this link, tag this object, tag this note.'
→
Don't just tag things. Use create_object to build a typed object that defines the relationship, and use add_tag only for broad categorization. This keeps the graph clean and structured.
Manual data transfer
Copying a web link and then having to manually paste the preview or metadata into a separate tool. This is slow and error-prone.
→
Use save_weblink. It handles the whole process: it saves the URL and automatically generates the dynamic preview and native metadata for you.
Searching by keyword only
Searching the whole space for 'meeting notes' and getting thousands of results, making it impossible to find the one specific object.
→
Don't just search. Use lookup to target a specific object hierarchy. You can narrow the search scope to find the exact node you're looking for.
When It Fits, When It Doesn't
Use this server if your primary goal is modeling complex relationships in a structured, object-based way. If you're building a personal knowledge base where 'how' things connect matters more than 'what' they are, this is for you. You're building a graph, not a filing cabinet.
Don't use this if you only need a simple chat bot to answer questions based on unstructured text, or if your data is mostly flat, unstructured documents. For that, a standard RAG pipeline might suffice. You need the object model and the ability to run tools like create_object to truly utilize it.
Independent Platform Disclaimer: Vinkius is an independent platform and is not affiliated with, endorsed by, sponsored by, verified by, or otherwise authorized by Capacities. 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 10 capabilities that interface natively with Claude, ChatGPT, Cursor, and any MCP client. No middleware. No custom integration required.
Available Capabilities
Tracking down a piece of information shouldn't require opening five different tabs.
Today, you find a great source online. You open a tab, copy the URL, and then open your knowledge management tool. You paste the link, but you have to manually copy the title and the summary, and then you have to paste that into a specific 'Weblink' field, hoping you didn't miss a crucial piece of metadata.
With the Capacities MCP Server, you just tell your agent: 'Save this link to my Research space.' The agent uses `save_weblink` and handles the whole process: it saves the URL, generates the preview, and attaches the metadata automatically. You get the link, fully baked, instantly.
Capacities MCP Server: Structured Data Modeling
Instead of writing everything as a long, unstructured block of text, you can tell your agent to 'Create a new Project object.' The agent uses `create_object` to instantiate a typed graph object. This enforces that the data you input has specific fields—a status, a date, a project owner—preventing you from leaving out key details.
The difference is control. You're not just dumping data; you're building a reliable system of interconnected, defined blocks. The data is always structured, no matter who is contributing.
Common Questions About Capacities MCP
How do I use the `get_space_info` tool with Capacities MCP Server? +
The get_space_info tool pulls all metadata about a specific space. It tells you exactly what object types exist and what properties they require. Use this first if you're trying to figure out the structure of a space before building an object.
Can `create_object` build a record that isn't in the space? +
No. create_object must comply with the predefined graph rules of the target space. You can't force it to create a record that doesn't fit the space's structure.
What's the difference between `lookup` and `get_object`? +
Use lookup when you know the general topic but not the exact ID. It searches the content across the space. Use get_object when you know the specific ID and need all the data points from that single record.
How does `add_tag` work with the Capacities MCP Server? +
add_tag links related items using structural categories. It's better than just typing a tag because it maintains the relationship in the graph, making the connection searchable.
How do I use the `save_weblink` tool to ensure the automatic preview generates correctly? +
The save_weblink tool sends the URL and triggers Capacities' native backend to generate the preview. You just need to provide the full URL string. The preview populates automatically when the content loads in your space.
What happens if I try to `create_object` with a structure that doesn't exist in the space? +
The system validates the request against the space's predefined structure parameters. If the structure is invalid, the tool returns an error detailing which fields need correction. You must use get_structures first to confirm the schema.
Can I use `list_spaces` to find all my personal knowledge areas? +
Yes, list_spaces provides a complete list of all top-level containers in your account. This allows your AI agent to choose the correct context before performing any other action like lookup or create_object.
How does `save_to_daily_note` handle formatting when I append code blocks? +
The tool accepts Markdown payloads, so you can wrap code in standard Markdown fences (```). The resulting content is appended to the daily note as formatted text, keeping the code syntax intact.
Can my AI agent automatically funnel research links into my Capacities daily note? +
Yes. It can append raw Markdown to your Daily Note dynamically, or use the dedicated WebLink endpoint to natively scrape and save beautifully previewed URLs into your specified space. This gives you a fast workflow for daily link-hoarding right from your chat tools.
Will the agent know the specific object structures configured in my custom Capacities space? +
Absolutely. Before creating nodes, your agent pulls constraints using the get_structures capability. It reads the schemas, understands which properties are locked or required (e.g., Dates vs Checkboxes), and constructs the new object JSON mapping perfectly inside the space.
Is the agent capable of categorizing objects with new tags automatically? +
Yes. Once an object is created or fetched via lookup, the AI can employ the tag modification tool to link categorical boundaries accurately. It's a frictionless way to organize hundreds of loosely coupled notes into connected knowledge topics without any manual toggles.
Use it with your favorite AI tools
Connect this server to Cursor, Claude, VS Code, and more.
More in this category
Gmail
Manage your inbox from AI — read, search, organize, and reply to emails across your Gmail efficiently.
Traction Guest
Manage visitor operations via Traction Guest — list hosts, locations, invites, sign-ins, and group visits directly from any AI agent.
eduMe
Equip your AI agent to manage mobile training, track trainees, and monitor course completion via the eduMe API.
You might also like
BreezoMeter Air Quality & Pollen
Universal air quality intelligence — get real-time AQI, pollutants, and pollen data via AI.
Nimble CRM
Manage relationships and deals via Nimble CRM — track contacts, deals, and activities directly from your AI agent.
Smaily Alternative
Manage email marketing campaigns, subscribers, and automations directly through Smaily.