Skip to main content
The MCP server provides 24 tools organized by category. Your AI assistant picks the right tools automatically based on what you ask — you don’t need to call them directly.
Infrastructure creation (datasources, datamarts, workspaces) is done in the Databrain UI. The MCP server provides read and linking operations for infrastructure, and full CRUD for everything from data apps onward. The semantic layer is the exception — it is populated and maintained entirely via MCP tools.

​
Infrastructure — Exploration & Linking

5 tools for discovering and connecting your existing Databrain resources. These are typically the first tools called during any setup flow. See Embed a Dashboard for a complete walkthrough.
ToolDescriptionAuth
list_datasourcesList all connected datasources in your organizationService Token
discover_datasource_schemaInspect a datasource to see its tables and columnsService Token
list_datamartsList all datamartsService Token
list_workspacesList all workspaces with their datamart and datasource connectionsService Token
link_datamart_to_workspaceLink a datamart to an existing workspaceService Token
Lists all datasources connected to your Databrain organization. Returns datasource names, types, and connection status.Parameters: None required.Use when: Starting a new setup to understand what data is available.
Introspects a datasource to return its tables, columns, and column types. Useful for understanding the shape of your data before creating datamarts.Parameters:
  • datasourceName (required) — Name of the datasource to inspect
Use when: You need to see what tables and columns are available in a specific datasource.
Lists all datamarts in your organization. Returns datamart names, associated datasources, and table configurations.Parameters: None required.Use when: Choosing which datamart to use for querying or embedding.
Lists all workspaces, showing which datamarts and datasources each is connected to.Parameters: None required.Use when: Understanding your workspace structure or selecting a workspace for dashboard operations.

​
Data Apps & Tokens

5 tools for managing data apps and their API tokens. These are handled automatically during embed setup — see Embed a Dashboard.
ToolDescriptionAuth
list_data_appsList all data appsService Token
create_data_appCreate a new data appService Token
list_api_tokensList API tokens for a data appService Token
create_api_tokenCreate and activate a new API tokenService Token
rotate_api_tokenRotate an existing API tokenService Token
Lists all data apps in your organization. Data apps are the top-level containers for embeds.Parameters: None required.Use when: Starting embed setup or managing existing data apps.
Creates a new data app.Parameters:
  • name (required) — Name for the new data app
Use when: You need a new data app to organize a set of embeds.
Lists API tokens for a specified data app.Parameters:
  • dataAppName (required) — Name of the data app
Use when: Checking whether an API token exists before embed operations.
Creates a new API token for a data app and automatically activates it for subsequent embed operations.Parameters:
  • dataAppName (required) — Name of the data app
Use when: The assistant needs an API token for embed CRUD and no active token exists.
Rotates an existing API token, generating a new key while invalidating the old one.Parameters:
  • dataAppName (required) — Name of the data app
Use when: Security rotation or when a token may have been compromised.

​
Embeds

7 tools for creating, configuring, and customizing embedded dashboards. See workflows: Brand & Theme, Filters, Multi-Tenant.
ToolDescriptionAuth
create_embedCreate an embed (standard or multi-tenant with clientId)API Token
list_embedsList all embed configurationsAPI Token
get_embed_detailsGet full embed details by IDAPI Token
update_embedUpdate embed configuration (raw patch)API Token
customize_embed_themePresets, colors, fonts, cards, chart appearance, responsive layoutAPI Token
configure_embedUI options, filters (string/date/number/SQL), i18n, translationsAPI Token
diagnose_embed_healthDiagnose malformed embeds and validate runtime readinessAPI Token
Creates an embed configuration. Supports standard dashboard embeds and multi-tenant embeds with a clientId for row-level security.Parameters:
  • datamartName (required) — The datamart to use in accessSettings
  • clientId (optional) — Client identifier for multi-tenant embeds
  • templateDashboardId (optional) — Dashboard to use as a template
  • isImportMetrics (optional) — Import metrics from the template dashboard
  • body (optional) — Raw passthrough to the API for advanced use cases
Use when: Setting up a new embed for a dashboard.
Lists all embed configurations for the current data app.Parameters: None required.Use when: Reviewing existing embeds or finding an embed ID.
Returns full details for a specific embed including access settings, theme, options, and filters.Parameters:
  • embedId (required) — The embed ID to inspect
Use when: Reviewing current configuration before making changes.
Applies a raw patch to an embed configuration. This is the low-level update tool — use customize_embed_theme or configure_embed for guided updates.Parameters:
  • embedId (required) — The embed ID to update
  • body (required) — Patch object sent to the API
Use when: Advanced updates that aren’t covered by the guided tools.
Applies visual theming to an embed: theme presets, custom colors, font family, card styling (padding, border radius, shadows), chart appearance (tooltips, legends, axes), and responsive breakpoints with layout columns.Parameters:
  • embedId (required) — The embed ID to theme
  • preset (optional) — Theme preset name
  • colors (optional) — Custom color configuration
  • fonts (optional) — Font family and sizing
  • cards (optional) — Card styling (padding, borders, shadows)
  • charts (optional) — Chart appearance (tooltips, legends, axes)
  • responsive (optional) — Breakpoint and column configuration
Use when: Branding an embed to match your product’s look and feel.
Configures embed behavior: UI toggle options, filters (string, date, number, and SQL-based server-side filters), internationalization (language, calendar type, translations), and custom messages.Parameters:
  • embedId (required) — The embed ID to configure
  • options (optional) — UI toggle options
  • stringFilters, numberFilters, dateFilters, sqlFilters (optional) — Filter configurations
  • hideDashboardFilters (optional) — Dashboard filters to hide
  • language (optional) — UI language
  • calendarType (optional) — Calendar format
  • translations (optional) — Custom translation strings
  • body (optional) — Raw passthrough for advanced use
Use when: Configuring embed behavior, filters, or localization.
Runs diagnostic checks on an embed to detect malformed metadata (missing access settings, invalid references) and optionally validates runtime readiness when guest token info is provided.Parameters:
  • embedId (required) — The embed ID to diagnose
  • guestTokenBody (optional) — Guest token payload to validate
  • permissions (optional) — Permissions object to validate
Use when: An embed isn’t rendering correctly or after complex configuration changes.

​
Frontend

2 tools for generating guest tokens and frontend integration code. These are the final step of every embed workflow — see Framework Integration for detailed per-framework guides.
ToolDescriptionAuth
generate_guest_tokenGenerate a frontend guest token (server-side)API Token
generate_embed_codeGenerate integration code for your frameworkNone (local)
Generates a guest token for frontend use. Guest tokens are the authentication mechanism for embedded dashboards — they’re generated server-side and passed to the frontend component.Parameters:
  • clientId (required) — Client identifier (use "None" if no tenancy)
  • dataAppName (required) — Name of the data app
  • params (optional) — Additional parameters: allowedEmbeds, appFilters, dashboardAppFilters, hideDashboardFilters, userIdentifier, timezone
  • permissions (optional) — Permission flags for what end-users can do
  • theme (optional) — Theme overrides for this token
  • expiryTime (optional) — Token expiration time
  • body (optional) — Raw passthrough to the guest token API
Use when: Generating the authentication token needed to render an embedded dashboard.
Generates framework-specific integration code for embedding a dashboard. Supports React, Next.js, Vue, Angular, Svelte, and vanilla JavaScript.Parameters:
  • framework (required) — Target framework
  • embedId (required) — The embed ID
  • token (optional) — Guest token to include in the generated code
Use when: Getting copy-pasteable frontend code after embed setup is complete. This tool runs locally and does not call the Databrain API.

​
Query & Semantic Layer

3 tools for natural language querying and managing data documentation. See workflows: Query Data, Populate Semantic Layer.
ToolDescriptionAuth
ask_questionAsk a question in natural language — returns data + chart suggestionService Token
get_semantic_layerInspect semantic layer metadata and quality scoreService Token
update_semantic_layerAdd or update table/column descriptions, synonyms, example questionsService Token
Converts a natural language question to SQL, executes it against your data, and returns results with a chart type suggestion. Requires a populated semantic layer for best results.Parameters:
  • question (required) — The question to ask
  • datamartName (required) — Which datamart to query (use list_datamarts to find available datamarts)
  • sessionId (optional) — Session ID for follow-up questions with conversation context
  • timezone (optional) — Timezone for date calculations
Use when: Querying data without building a dashboard. Great for ad-hoc analysis and data exploration.
Retrieves semantic layer metadata for a datamart: table and column descriptions, synonyms, example questions, and a completion score indicating documentation quality.Parameters:
  • datamartName (required) — Name of the datamart
Use when: Checking whether the semantic layer is ready for natural language querying, or reviewing current documentation quality.
Adds or updates semantic layer metadata: table descriptions, column descriptions, synonyms, and example questions. This is required before ask_question produces accurate results.Parameters:
  • datamartName (required) — Name of the datamart
  • tables (optional) — Table-level descriptions and metadata
  • columns (optional) — Column-level descriptions and synonyms
  • exampleQuestions (optional) — Sample questions to guide the AI
  • feedback (optional) — General feedback for the semantic layer
Use when: Bootstrapping or improving the semantic layer for better query accuracy.

​
Discovery & Orchestration

2 tools for dashboard discovery and guided setup flows.
ToolDescriptionAuth
list_dashboardsList dashboards by data app, workspace, or globallyService Token / API Token
setup_embed_interactiveGuided step-by-step embed setup flowService Token + API Token
Lists dashboards scoped by data app (if API token is active), workspace name, or globally.Parameters:
  • workspaceName (optional) — Filter by workspace
  • dataAppId (optional) — Filter by data app
Use when: Finding dashboards to embed or review.
A guided, step-by-step flow that walks through the entire embed setup process. Each step returns a status and instructions, and the assistant passes the state forward through the flow.Steps: discover_data_apps → choose_data_app / create_data_app → ensure_api_token → discover_workspaces → choose_workspace → list_dashboards → choose_dashboard → create_embed → runtime_readiness → completeParameters:
  • step (required) — The current step in the flow
  • state (optional) — State object from the previous step (pass forward unchanged)
  • Various step-specific parameters
Use when: The assistant orchestrates the full embed setup from start to finish.