Skip to main content
Prompts are guided workflows that your AI assistant can follow. They encode best practices, safe sequencing, and the correct tool order for common tasks. You can trigger a prompt by asking naturally or by referencing the prompt name directly.

​
At a Glance

PromptCategoryWhat It Does
embed-existing-dashboardSetupEmbed a dashboard that already exists in the Databrain UI
embed-blank-dashboardSetupCreate per-client dashboard embeds for multi-tenant apps
brand-my-embedCustomizationCustomize colors, fonts, cards, charts, responsive layout, and access presets
query-dataData explorationQuery a datamart with natural language and use the semantic layer
populate-semantic-layerData setupGenerate and save table/column descriptions, synonyms, and examples
customize-customer-dashboardDashboard updatesAdd, update, remove, or generate widgets for one customer’s dashboard
sql-to-metric-migrationProduction migrationValidate SQL, dry-run metric creation/update, then apply with explicit confirmation

​
Setup Prompts

Embed a dashboard you’ve already built in the Databrain UI.The most common starting point. Walks through data app selection, API token creation, workspace and dashboard selection, embed creation or reuse, guest token generation, and frontend code generation.Example asks:
  • “Embed my sales dashboard in my React app”
  • “Set up a Databrain embed for my dashboard”
  • “I want to embed the revenue overview dashboard”
Tools used: list_data_apps -> list_api_tokens / create_api_token -> list_workspaces -> list_dashboards -> list_embeds -> create_embed -> get_embed_details -> generate_guest_token -> generate_embed_codeSee full workflow ->
Multi-tenant setup where each client gets their own dashboard.Creates per-client dashboard embeds using clientId for customer-specific dashboard copies. This is the right flow when customers need different layouts, self-serve metrics, or per-client dashboard isolation.Example asks:
  • “Create per-client dashboard embeds for my SaaS app”
  • “Set up multi-tenant embeds with separate dashboards per customer”
  • “I need each client to have their own dashboard copy”
Tools used: list_data_apps -> list_api_tokens / create_api_token -> list_dashboards -> list_datamarts -> create_embed per client -> generate_guest_token -> generate_embed_codeSee full workflow ->

​
Data Prompts

Query your data with natural language and manage semantic context.Ask questions about your data without setting up dashboards. The assistant verifies the datamart and semantic layer before calling ask_question.Example asks:
  • “What was total revenue last month?”
  • “Compare sales across regions for Q1”
  • “Show me the top 10 products by revenue”
Tools used: list_datamarts -> get_semantic_layer -> update_semantic_layer or start_semantic_layer_generation if needed -> ask_questionSee full workflow ->
Generate and save table/column descriptions to improve natural language querying.Bootstraps the semantic layer by generating business-friendly descriptions, synonyms, and example questions, then pushes them to Databrain.Example asks:
  • “Set up the semantic layer for my orders datamart”
  • “Add descriptions to all my tables and columns”
  • “Improve the semantic layer quality for better query accuracy”
Tools used: list_datamarts -> get_semantic_layer -> update_semantic_layer -> get_semantic_layer -> ask_questionSee full workflow ->

​
Embed Customization Prompt

Customize colors, fonts, chart styles, responsive layout, cards, and access presets.Applies visual theming to match your product’s branding. Supports theme presets, custom colors, font family, card styling, chart appearance, responsive breakpoints, and access presets.Example asks:
  • “Make my embed match our brand: primary color #4F46E5, dark mode”
  • “Apply the corporate theme and use Inter font”
  • “Customize chart colors to use our brand palette”
Tools used: list_embeds -> get_embed_details -> customize_embed_theme -> update_embed when advanced patching is neededSee full workflow ->

​
Customer Dashboard Prompts

Add, modify, remove, or generate widgets for one customer’s dashboard.This flow is designed for customer-specific dashboard updates. The assistant asks for clientId and dashboardName, resolves the required resources automatically, and keeps every change scoped to one customer.Example asks:
  • “Add the Calls per Day widget to Acme’s dashboard”
  • “Remove the billing widget for one client only”
  • “Generate a widget showing disqualification reasons for this customer”
  • “Give me a demo link so I can see what this client sees”
Tools used: list_tenants when clientId is unknown -> list_widgets -> generate_widget or add_widget -> update_widget / remove_widget -> list_widgets -> get_demo_linkSee full workflow ->
Production-safe SQL-to-metric migration.Converts validated warehouse SQL into a Databrain metric/widget with an explicit validation, dry-run, review, apply, and verification sequence.Example asks:
  • “Validate this SQL as a new metric on my operations dashboard”
  • “Dry-run creating this converted query as a metric, then show me the plan”
  • “Apply the approved metric migration”
Tools used: list_workspaces -> list_dashboards -> list_widgets -> validate_metric_spec -> create_or_update_metric_from_query in dry_run -> create_or_update_metric_from_query in apply with confirm: "APPLY_TO_PRODUCTION" -> list_widgets -> get_demo_linkSee full workflow ->

​
Prompt Behavior

  • Prompt flows are interactive. The assistant presents results and waits for user input before the next mutating step.
  • Token and resource details stay in your MCP configuration and Databrain workspace. The assistant does not need service tokens, API tokens, data app IDs, datasource IDs, or embed IDs unless you provide them proactively.
  • Use list_tenants to discover valid clientId values from the workspace tenancy database before guest-token, demo-link, scheduled-report, or customer-dashboard widget work.
  • Use externalMetricId, not metricId, for update_widget and remove_widget.
  • create_datamart, create_workspace_metric, and create_or_update_metric_from_query apply operations require confirm: "APPLY_TO_PRODUCTION".

Workflows

See these prompts in action with end-to-end walkthroughs

Tools Reference

Detailed docs for every tool