Ninox MCP. Run database operations from natural conversation.
Gemini Works with every AI agent you already use
…and any MCP-compatible client
Just plug in your AI agents and start using Vinkius.
Ninox MCP Server connects your AI agent directly to a low-code database platform. It lets you manage structured data—read records, build new entries, run complex scripts, and check schemas—all through natural conversation.
Stop using spreadsheets for mission-critical business logic; this gives your AI the power of a dedicated database administrator.
What your AI agents can do
Create records
Adds brand new records or updates data when a specific field needs to be populated.
Delete record
Removes an existing record permanently from a specified Ninox table.
Execute ninox script
Runs advanced, native Ninox scripts or complex queries to perform custom calculations or data manipulation.
List all databases in your team or retrieve the detailed field schema for a specific table to understand its data layout.
List multiple records from a table, or pull specific details for one record when you know its ID.
Add brand new entries to any Ninox table using structured data input through the AI interface.
Update specific fields on records that already exist, ensuring your business metrics stay current.
Permanently remove records from a table when they are outdated or incorrect.
Execute advanced, native Ninox scripts and complex queries to perform custom calculations or data transformations.
Ask AI about this MCP
Supported MCP Clients
OAuth 2.0 Compatible Gemini Waiting for input…
Ninox MCP Server: 8 Tools for Database Ops
Use these eight tools to read, write, modify, and manage every aspect of your structured Ninox database directly through your AI agent.
Make your AI actually useful.
Add this MCP to Claude, Cursor, or Windsurf and your AI stops guessing. It gets real tools to look things up, take action, and handle the stuff you keep doing by hand.
Start using Ninox on Vinkius019dd12dcreate records
Adds brand new records or updates data when a specific field needs to be populated.
019dd12ddelete record
Removes an existing record permanently from a specified Ninox table.
019dd12dexecute ninox script
Runs advanced, native Ninox scripts or complex queries to perform custom calculations or data manipulation.
019dd12dget database schema
Retrieves the full field definition and structure for any database table you reference.
019dd12dget record
Pulls all details for a single, specific record when provided with its unique ID.
019dd12dlist databases
Lists every database and workspace your team manages within Ninox.
019dd12dlist records
Retrieves a list of records from an entire table, allowing you to see the top-level data points across multiple entries.
019dd12dupdate record
Modifies specific fields on existing records without deleting and recreating them.
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 Ninox, then connect any of our 5,000+ other servers whenever your AI needs more. One click, no limits.
- Use this MCP plus 5,000+ 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
Independent Platform Disclaimer: Vinkius is an independent platform and is not affiliated with, endorsed by, sponsored by, verified by, or otherwise authorized by Ninox. 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 8 capabilities that interface natively with Claude, ChatGPT, Cursor, and any MCP client. No middleware. No custom integration required.
Finding core business metrics shouldn't require navigating five different tabs.
Right now, getting a full picture of an account means logging into Ninox. You navigate to the 'Customers' database, find the correct table, then click on 'Orders,' which takes you to another view. If you need data from three different related tables—like customer status, product list, and invoice date—you have to manually cross-reference fields, copy IDs, and stitch everything together in a spreadsheet.
With this MCP server, the process is simple: you just ask your agent. 'Show me all active customers who haven't placed an order since last month.' The agent handles the multi-table join, runs the necessary tools (`list_records`, `get_database_schema`), and returns one clean answer in chat.
Using Ninox MCP Server: Full Control with execute_ninox_script
Manually running complex reports or performing data transformations means you have to write custom scripts, deal with syntax errors, and figure out the exact query language for joins across multiple tables. It's slow, brittle, and requires dedicated developer time.
Now, you tell your agent what needs to happen—'Calculate the total profit margin for all inventory items flagged as low stock.' The agent handles the complexity by executing the native Ninox script via `execute_ninox_script`. You get the result immediately. It’s done.
What you can do with this MCP connector
The Ninox MCP Server plugs your AI agent directly into a structured, low-code database platform. You'll manage mission-critical data—reading records, building new entries, running complex scripts, and checking schemas—all through natural conversation with your agent. Stop using messy spreadsheets for anything that matters; this gives your AI the full control of a dedicated database administrator.
Discovering Your Data Structure
You can start by asking your agent to list every database or workspace your team manages across Ninox. Need to see how a specific table is set up? You'll use get_database_schema to retrieve the complete field definition and structure for any table you reference, letting you understand exactly what data points are available.
Viewing Records
Your agent can pull an overview of multiple entries from an entire table using list_records, giving you a quick look at all the top-level data points. If you know the unique identifier for a specific record, you'll use get_record to pull every single detail associated with that one entry.
Modifying Data Entries (CRUD Operations)
To add new information, your agent uses create_records, which builds brand-new records or populates data when a specific field needs input. If the record already exists but some details are wrong, you'll use update_record to modify only the necessary fields without deleting and recreating the whole thing. You can permanently wipe out outdated information using delete_record, removing an existing record from any specified table.
Running Custom Logic
For complex calculations or specialized business rules that go beyond simple data entry, you'll run advanced, native Ninox scripts and complex queries directly with execute_ninox_script. This mechanism lets your agent perform custom data transformations or execute intricate logic the way a database administrator would. It’s full database power, all through natural language conversation.
How Your Agent Interacts With Everything
Your AI client uses these tools to handle the entire lifecycle of structured business information. You can first check what databases exist using list_databases, then get a schema definition with get_database_schema. When you need to write new data, your agent calls create_records and populates fields; when records are stale, it deletes them via delete_record; if the record is just wrong, it updates specific fields using update_record.
To read information, it can list multiple entries with list_records or pull all details on a single ID using get_record, and for maximum complexity, you'll execute custom business logic through execute_ninox_script.
019dd12e-1454-726e-9ec0-eac613c1ed6c 
How Ninox MCP Works
- 1 First, subscribe to the server on Vinkius Marketplace and enter your Ninox Personal Access Token and Team ID.
- 2 Next, reference a specific database action in your chat prompt—for example, 'Show me the schema for the Customers table.'
- 3 Finally, your AI client executes the required tool (like
get_database_schema), pulls the data, and reports the results back to you in plain text.
The bottom line is: You talk naturally, the agent runs the necessary Ninox tools, and you get structured data returned instantly.
Who Is Ninox MCP For?
This is for people who spend too much time jumping between dashboards, spreadsheets, and database interfaces. If your job involves checking project status, running metrics reports, or managing core business records that shouldn't live in Excel, you need this. It gives the AI a direct line to your source of truth.
You use it to quickly retrieve project statuses across different tables or check team access using simple commands without logging into Ninox.
You automate the retrieval of large sets of table records (list_records) and execute custom scripts via natural conversation, turning raw data into actionable insights.
You integrate real-time database management—from running complex queries with execute_ninox_script to managing the full record lifecycle—right within your chat environment.
What Changes When You Connect
- Real-time Data Checks: Instead of opening a table and manually filtering, you ask your agent to run
list_recordsorget_record. You get the exact data point (like 'Project Status' or 'Revenue') instantly, without clicking through dashboards. - Full Lifecycle Control: The ability to
create_records,update_record, anddelete_recordmeans your AI isn't just reading data; it's actively managing your business truth. This keeps records accurate and current right from the chat window. - Schema Clarity: Need to know what fields are available before writing a query? Use
get_database_schema. It pulls the table definition, so you never guess what column name or data type is correct again. - Complex Logic on Demand: The
execute_ninox_scripttool lets you run advanced business logic—think complex joins or custom calculations—that would normally require a dedicated scripting environment. You just ask for it. - Instant Context Switching: No more switching from your chat client to Ninox, navigating menus, and copying IDs. The agent handles all the context routing, keeping you in one place while running multi-step operations.
Real-World Use Cases
Checking a Project Status Across Teams
The Ops Lead needs to know if Acme Corp's invoice is ready. They ask the agent, 'What's the status of Invoice 1847?' The agent uses get_record on the Invoices table and returns the exact status and due date immediately. No dashboard clicks needed.
Onboarding a New Client Record
A new customer walks in. Instead of filling out a multi-tab form, the agent uses create_records to populate the Customer table with all necessary details (Name, Email, Company). The record is created and linked instantly.
Auditing Old Data
A data analyst needs to see every customer who signed up last quarter but hasn't placed an order. They prompt the agent to 'list records from Customers where Status is Lead.' The agent executes list_records and gives them a clean, filtered list.
Running Quarterly Financial Reports
The manager needs a complex calculation: total revenue minus all pending payments. Instead of writing SQL or running a custom script manually, they ask the agent to 'run the quarterly net revenue script.' The agent uses execute_ninox_script and delivers the final number.
The Tradeoffs
Assuming Data Availability
The user tries to ask, 'What are all my sales records?' without knowing which table holds them. They might just start listing fields and get an error.
→
First, run list_databases to see your available structures. Then, use get_database_schema on the correct database (e.g., 'Sales') before attempting to list or query records.
Over-relying on Manual Updates
The user manually copies a record ID and then types, 'Change this amount.' They forget which table the ID belongs to.
→
Always use get_record first. Specify the database and table context in your prompt before asking for an update. This ensures the agent targets the correct instance using update_record.
Trying to Build Logic with Chats
The user asks, 'If this happens, then do that.' They treat the chat like a programming environment.
→
For multi-step logic or calculations, you must use execute_ninox_script. This tool runs code directly against Ninox's engine; it doesn't just talk about doing it.
When It Fits, When It Doesn't
Use this server if your core business processes rely on highly structured, relational data (i.e., customer records linked to invoices linked to products). If you need to perform CRUD operations or run complex queries that require knowing a specific schema, this is the tool. Don't use it if you just need a simple chat log or unstructured text storage; those are better handled by general messaging APIs.
Crucially: Never try to bypass get_database_schema. If your agent fails on a query, the first thing to check is whether you have retrieved the schema for that table. This prevents invalid state changes and tells you exactly what fields exist before you write code or data.
Common Questions About Ninox MCP
How do I use the create_records tool in Ninox? +
You simply prompt your agent with the data and context, for example: 'Create a new record in the Orders table for Acme Corp, amount $4500.' The agent handles mapping that natural language request to the structured call using create_records.
What is the difference between list_records and get_record with Ninox? +
list_records shows you a summary view of many entries in an entire table. Use it when you need to see if records exist or check general status. get_record pulls every single detail for one specific entry, requiring the record's unique ID.
Can I run complex queries using execute_ninox_script? +
Yes. The execute_ninox_script tool lets you run native Ninox scripts and advanced logic directly. This is how you perform calculations or data manipulations that simple CRUD tools can't handle.
If I need to check what fields a table has, which tool do I use? (get_database_schema) +
You must use get_database_schema. This tool reads the metadata for a given table and returns its field names, data types, and whether they are required. It's your blueprint.
Before I use list_databases or any other tool, what setup is required? +
You need a Ninox Personal Access Token and Team ID. Your AI client uses these credentials to authorize every call, ensuring you only manage databases within your specific team structure.
If I use the delete_record tool and the record is linked elsewhere, what happens? +
The deletion process respects data dependencies defined in Ninox. It checks for related records before confirming a delete; if constraints exist, it alerts you instead of failing silently.
How does update_record handle changing just one field on a record? +
You specify the unique record ID and only the fields you want to change. The server modifies those exact parameters, preventing unintended overwrites on other data points in that record.
What does list_databases show about my team’s overall structure? +
It returns a complete inventory of every managed database name and ID under your team's scope. This lets you quickly identify all potential data sources before targeting any specific table or record for action.
Can my AI automatically find my Team ID for me? +
The Team ID is required during initial setup. You can find it in your browser URL when you are inside your Ninox workspace (e.g., app.ninox.com/team/ABC123XYZ).
How do I find my Ninox Personal Access Token? +
Log in to your Ninox account at app.ninox.com, click the gear icon for Settings, and navigate to the API section to generate your unique secret token.
Can I run native Ninox scripts via the AI? +
Yes! Use the execute_ninox_script tool. Provide your database ID and the script body, and the AI will perform the operation using Ninox's native server-side scripting engine.
Use it with your favorite AI tools
Connect this server to Cursor, Claude, VS Code, and more.