# Databox MCP MCP > Databox helps you manage and analyze business intelligence metrics directly through natural conversation. Connect any AI agent to Databox via Vinkius to read data schemas, push raw records into datasets, and check API activity logs for real-time reporting. ## Overview - **Category:** data-analytics - **Price:** Free - **Tags:** kpi-tracking, data-visualization, real-time-dashboards, data-ingestion, performance-metrics, business-analytics ## Description Stop manually querying databases just to update a metric or check usage quotas. This MCP lets your AI client take over your entire data ingestion workflow. You talk to it—'What were the sales numbers last week?'—and it handles the complex steps: first, it finds the correct dataset; second, it runs the necessary checks on the schema and current storage limits; finally, it pulls the clean metrics you need for instant visualization. It’s like having a dedicated data engineer sitting at your desk who answers questions in plain English. Vinkius hosts this connection so your AI agent can access all of Databox's tools from one place. ## Tools ### create_data_source Creates and registers a connection to an external data source within Databox. ### create_dataset Initializes a brand new, empty dataset collection in the system. ### delete_dataset Permanently removes an existing dataset and all its associated data. ### get_dataset_details Fetches specific schema information, including primary keys, for a named dataset. ### get_current_user Retrieves and reports the authenticated profile details for the connected user account. ### get_storage_statistics Reports current data storage usage and consumption metrics for the account. ### list_accounts Lists all Databox accounts connected or managed under the primary API key. ### list_data_sources Displays a directory of all active and configured external data source integrations. ### list_datasets Retrieves a comprehensive list of every available dataset collection in the account. ### list_activity_logs Retrieves a chronological record of recent API interactions and data operations performed. ### list_dataset_metrics Lists all defined metrics, fields, and column types within a specific dataset. ### push_metrics_data Ingests structured, raw data records directly into a specified dataset for immediate visualization. ## Prompt Examples **Prompt:** ``` List all datasets in my Databox account. ``` **Response:** ``` I've retrieved your datasets. You currently have 3 collections: 'Daily Sales' (ID: ds_123), 'User Engagement', and 'API Logs'. Which one would you like to push data to? ``` **Prompt:** ``` Push record to 'ds_123': value 1500, date '2026-04-16'. ``` **Response:** ``` Data pushed successfully! I've ingested the value 1500 for 2026-04-16 into the 'Daily Sales' dataset. The metric is now being processed for your dashboards. ``` **Prompt:** ``` Show my storage usage and API activity logs. ``` **Response:** ``` Retrieving metrics... You're using 45% of your data storage quota. Recent logs show 10 successful pushes to 'Daily Sales' in the last hour. Everything is running smoothly. ``` ## Capabilities ### Manage data collections Your agent lists, retrieves schema details for, and deletes entire database collections (tables) within Databox. ### Ingest raw metrics The MCP allows pushing arrays of new, raw data records directly into a specified dataset to update real-time dashboards. ### Track usage and connectivity You can check current storage consumption statistics, review authenticated user profiles, and access API activity logs for auditing purposes. ## Use Cases ### Need to update yesterday's sales data quickly The Ops Lead needs to record 50 new transactions that haven't hit the main database yet. The agent handles this by first listing the correct dataset (`list_datasets`) and then executing `push_metrics_data` with the raw transaction records. ### Auditing data access for compliance The Security Analyst needs to prove who accessed which datasets yesterday. They ask the agent, and it runs `list_activity_logs`, providing a detailed record of successful pushes or schema reads. ### Discovering all available metrics A new Data Scientist joins the project. Instead of reading documentation, they ask the agent to run `list_dataset_metrics` on 'User Engagement' and instantly get a list of every trackable field. ### Setting up a new data feed The team needs metrics from a newly acquired source. They use `create_data_source` to link the external API, followed by `list_data_sources` to confirm the connection is active and working. ## Benefits - You bypass manual data entry. Instead of copying raw CSVs, you can tell the agent to push new records using `push_metrics_data`, keeping dashboards current instantly. - Stay within budget by checking quotas automatically. Use `get_storage_statistics` or view API activity logs via `list_activity_logs` before running major data pushes. - Schema guesswork is gone. Before you build anything, ask the agent to run `get_dataset_details` on a dataset name and get the exact structure back. - Manage your entire infrastructure in one chat session. You can list all connected accounts with `list_accounts` and audit which data sources are active with `list_data_sources`. - The AI handles complexity by allowing you to query metadata like available metrics using `list_dataset_metrics` without writing a single SQL line. ## How It Works The bottom line is that your AI client acts as a universal interface for Databox, translating conversational requests into specific data commands. 1. Subscribe to this MCP on Vinkius. Then, grab your Databox API Key (v1) from the dashboard settings. 2. Connect your preferred AI client and prompt it with a data task—for example, asking it to list all available datasets or push new sales figures. 3. Your agent uses the appropriate tool to perform the action, giving you instant confirmation of success, status, or required input parameters. ## Frequently Asked Questions **How do I check my data usage with Databox? (using get_storage_statistics)** Run `get_storage_statistics` in your agent. It immediately reports how much storage you've used and what the remaining quota is, so you can plan major data ingestion jobs. **What if I need to change a dataset name? (using delete_dataset)** You first list all available datasets using `list_datasets` to confirm the exact ID. Then you use `delete_dataset` to remove it, making sure you have backups before running that command. **Can I see who has accessed my data? (using list_activity_logs)** Yes. The agent runs `list_activity_logs`, providing a detailed log of every API call and data operation performed on your account for auditing purposes. **What is the difference between listing datasets and listing metrics? (using list_datasets vs list_dataset_metrics)** Using `list_datasets` shows you the container names (the tables). Using `list_dataset_metrics` tells you the specific columns, fields, and types *inside* one of those containers. **How can I check my profile and permissions using get_current_user?** The tool returns your authenticated user details. This confirms which account the AI agent is operating under, ensuring data security before making changes or running reports. **When should I use create_data_source?** You run this when you need to establish connection credentials for an external system. It initializes the link so your agent can start pulling metrics from that brand-new source into Databox. **Using get_dataset_details, what metadata can I retrieve about a specific collection?** It provides deep structural information, including schema definitions and primary key setups. This is essential for understanding exactly how your data is organized before you push metrics. **If I need to ingest new records, how do I use push_metrics_data?** You pass an array of raw data records along with the target dataset ID. The agent pushes this batch directly into Databox for real-time visualization and immediate metric processing. **How do I find my Databox API Key?** Log in to your account, navigate to **Account Settings** > **API Tokens**, and copy your unique v1 API Key. **Can I create new datasets via AI?** Yes! Use the `create_dataset` tool. You'll need to specify a title, a source ID, and an array of primary keys for the table structure. **Does it support real-time data pushing?** The `push_metrics_data` tool allows for immediate ingestion of data records, making them available for visualization in Databox instantly.