Skip to main content

Use MCP in chat-based AI tools Private Preview

You can use Atlan MCP (Model Context Protocol) to access Atlan’s metadata, lineage, and glossary directly from chat-based AI tools. It enables teams to explore, query, and manage data assets directly within their chat environment.

  • Search and discover assets: Find datasets, dashboards, and models using metadata filters.
  • Explore lineage: Understand data dependencies, upstream sources, and downstream impact.
  • Query metadata and data: Use Atlan’s DSL to generate and execute optimized queries.
  • Update metadata: Certify assets, enrich descriptions, or align documentation.
  • Manage glossaries: Create, organize, and maintain business terms and categories at scale.

This page provides examples of how to use Atlan MCP across tools like Claude, Cursor, Copilot, and ChatGPT to perform common data discovery, lineage analysis, and glossary management tasks.

info

Make sure Atlan MCP is configured and the required tools (for example, Search Assets, Explore Lineage, or Glossary) are enabled before running any prompts. For setup, see Set up Atlan MCP.

Data discovery​

Use these prompts to find, analyze, and visualize data assets within your chat-based AI tool.
Data discovery helps you search across Atlan, query datasets, and understand lineage and relationships between assets.

Assess lineage and downstream impact​

Use this to understand how changes to a dataset affect dashboards, reports, or downstream assets.
It helps evaluate dependencies before applying schema or logic changes.

To assess lineage and downstream impact, you need these Atlan MCP tools:

  • Explore lineage: to trace downstream assets and their dependencies
  • Search assets: to locate impacted reports, dashboards, and models in your catalog

Examples: If you want to check which dashboards depend on a table, use this prompt in your chat tool:

For the table `orders_summary`, list all downstream assets and identify impacted dashboards.

This retrieves the full list of affected dashboards, usage information, and ownership details.

Review SQL changes with lineage context​

Use this to assess the lineage impact of SQL changes during pull requests or code reviews.

To review SQL changes with lineage context, you need these Atlan MCP tools:

  • Explore lineage: to evaluate downstream dependencies and impact scope
  • Query by DSL: to identify and analyze impacted assets through targeted queries

Examples: When reviewing a pull request, use this prompt in your chat tool:

Analyze this SQL diff for potential downstream impact using Atlan MCP.
Return a summary with impacted assets, risk level, and owners.

This identifies affected assets, risk levels, and owners before a change is merged.

Analyze data lineage and impact​

Use this to trace asset dependencies, identify impacted objects, and perform impact analysis.

To analyze data lineage and impact, you need these Atlan MCP tools:

  • Explore lineage: to visualize upstream and downstream relationships between data assets

Examples: To list downstream assets for a table, use this prompt in your chat tool:

Get all downstream assets for the table CUSTOMER_TRANSACTIONS.

To view immediate upstream lineage, use this prompt in your chat tool:

Show the immediate upstream lineage for default/snowflake/123456/analytics/orders.

To retrieve lineage up to three levels, use this prompt in your chat tool:

Retrieve full upstream lineage (depth 3) for the table SALES_SUMMARY.

Asset management​

Use these prompts to maintain and enrich asset metadata, certifications, and documentation in Atlan.
Asset management tasks help maintain data quality, consistency, and governance across the catalog.

Search assets by certification and popularity​

Use this to find assets by certification status, popularity, or usage frequency.
This helps identify trusted or frequently used assets.

To search assets by certification and popularity, you need these Atlan MCP tools:

  • Search assets: to filter assets by certification status, popularity scores, and usage metrics

Examples: To find verified, popular assets, use this prompt in your chat tool:

Find the top 10 high-popularity tables related to customers that are certified.

To find uncertified but frequently accessed assets, use this prompt in your chat tool:

Find frequently accessed but uncertified tables from the Snowflake connection default/snowflake/123456.

Update and enrich asset metadata​

Use this to update asset certifications, add user descriptions, or modify README documentation.
This keeps metadata accurate and consistent.

To update and enrich asset metadata, you need these Atlan MCP tools:

  • Update assets: to modify asset certifications, user descriptions, and README documentation

Examples: To certify an asset, use this prompt in your chat tool:

Mark the table CUSTOMER_TRANSACTIONS as certified (VERIFIED).

To add a user description, use this prompt in your chat tool:

Add a user description "Contains customer purchase history" to the asset default/snowflake/123456/sales/customer_transactions.

To update the README content, use this prompt in your chat tool:

Update the readme of the asset xyz.

Discover top-performing and uncertified assets​

Use this to identify active or uncertified assets for governance prioritization.
It supports curation and certification planning.

To discover top-performing and uncertified assets, you need these Atlan MCP tools:

  • Search assets: to filter assets by activity levels, certification status, and popularity metrics

Examples: To find active and verified assets, use this prompt in your chat tool:

Retrieve the top 20 active verified assets across all data sources by popularity score.

To find uncertified active assets, use this prompt in your chat tool:

Filter assets that are active but have no certificate status assigned.

Data quality rules​

Use these prompts to create, adjust, schedule, and delete data quality rules on your tables and views directly from chat-based AI tools.
These workflows help you operationalize data quality checks alongside your development and analytics work.

Create initial data quality rules​

Use this to define null checks, row count checks, freshness checks, or custom SQL-based rules on critical assets.
This helps you establish a baseline set of quality controls on key tables and columns.

To create initial data quality rules, you need these Atlan MCP tools:

  • Create DQ rules: to define new data quality rules on specific tables, views, materialised views, or Snowflake dynamic tables

Examples: To create core quality rules on important tables, use prompts like:

Create a null check rule on the EMAIL column in the CUSTOMERS table. The null count should be less than 5 and mark it as URGENT priority.
Set up a row count rule for the ORDERS table that alerts if there are fewer than 1000 rows.
Add a freshness check on the LAST_UPDATED column in the USERS table. Alert if the data is older than 2 days.
Create a custom SQL rule on the TRANSACTIONS table that alerts if any REVENUE values are negative or exceed 1,000,000.

Adjust rule thresholds and priorities​

Use this to update thresholds, priorities, and conditions for existing data quality rules as data patterns evolve.
This helps you calibrate sensitivity to reduce false positives while still catching real issues.

To adjust rule thresholds and priorities, you need these Atlan MCP tools:

  • Update DQ rules: to modify thresholds, alert priorities, and rule conditions for existing quality rules

Examples: To tune existing rules, use prompts like:

Update the null count threshold for the EMAIL column rule on the CUSTOMERS table to 50 instead of 5.
Change the priority of the row count rule on the ORDERS table from NORMAL to URGENT.
Modify the freshness rule on the USERS table to alert when data is older than 3 days instead of 2.
Update all duplicate count rules on the CUSTOMERS table to use a threshold of 10.

Schedule quality checks during ETL windows​

Use this to schedule data quality rules to run automatically using cron expressions.
This helps you align quality checks with ETL jobs or off-peak windows.

To schedule quality checks during ETL windows, you need these Atlan MCP tools:

  • Schedule DQ rules: to configure cron-based schedules for rules on tables, views, materialised views, and Snowflake dynamic tables

Examples: To schedule rules around ETL workflows, use prompts like:

Schedule all data quality rules on the CUSTOMERS table to run daily at 2 AM UTC.
Set up automatic quality checks for the ORDERS table to run every 6 hours.
Configure weekly quality rule execution for the ANALYTICS_SUMMARY view every Sunday at midnight UTC.
Schedule quality checks on the SALES materialised view to run daily at 3 AM America/New_York time.

Clean up deprecated or replaced rules​

Use this to delete data quality rules that are outdated, noisy, or replaced by better logic.
This helps you keep your rule set lean and focused on what matters.

To clean up deprecated or replaced rules, you need these Atlan MCP tools:

  • Delete DQ rules: to remove one or more data quality rules by GUID

Examples: To remove outdated rules, use prompts like:

Delete the email validation quality rule with GUID abc-123-def.
Delete all data quality rules that were created for testing last week on the CUSTOMERS table.
Delete the data quality rule that is causing false alerts on the STATUS column in the ORDERS table.

Business glossary​

Use these prompts to create, explore, and manage business glossary terms and categories.
Glossary actions help maintain consistent business definitions and align data with documentation.

Explore glossary relationships​

Use this to explore business glossary terms, their relationships, and linked datasets.
It helps connect business concepts to actual data.

To explore glossary relationships, you need these Atlan MCP tools:

  • Glossary: to fetch business terms, definitions, and their relationships
  • Search assets: to map linked datasets, dashboards, and other assets to glossary terms

Examples: To explore terms related to customer churn, use this prompt in your chat tool:

Find glossary terms related to "Customer Churn" and show how they connect to datasets or dashboards.

This returns related glossary terms with definitions, ownership, and associated datasets.

Align glossary with documentation​

Use this to align glossary definitions with documentation in sources such as Confluence.
It ensures business terms remain consistent across tools.

To align glossary with documentation, you need these Atlan MCP tools:

  • Search assets: to locate glossary terms and their current definitions
  • Update assets: to modify or add references and documentation links
  • Glossary: to maintain and update business term definitions

Examples: To check and sync definitions, use this prompt in your chat tool:

Check if definitions from the Confluence "Customer Metrics" page align with glossary terms in Atlan.
Update Atlan glossary if discrepancies are found.

Create glossaries, categories, and terms​

Use this to define new glossaries, categories, and business terms.
This helps teams maintain shared understanding and governance.

To create glossaries, categories, and terms, you need these Atlan MCP tools:

  • Create glossaries: to define new business glossaries with descriptions and metadata
  • Create glossary categories: to organize related terms within glossaries
  • Create glossary terms: to define and manage individual business concepts

Examples: To create a glossary, use this prompt in your chat tool:

Create a new glossary named "Customer Data Glossary" with description "Business definitions for customer-related entities."

To add a category, use this prompt in your chat tool:

Create a category "Customer Demographics" under the "Customer Data Glossary".

To add a verified term, use this prompt in your chat tool:

Create a verified term "Customer Lifetime Value (CLV)" under the "Customer Data Glossary".

Bulk glossary management​

Use this to create glossaries, categories, and terms in bulk from spreadsheets, dataframes, or documentation.
This accelerates metadata onboarding and ensures consistency.

To perform bulk glossary management, you need these Atlan MCP tools:

  • Create glossaries: to define multiple glossaries in one operation with consistent metadata
  • Create glossary categories: to add categories at scale across multiple glossaries
  • Create glossary terms: to bulk-create or update business terms from external sources

Examples: To bulk-create glossaries and categories, use this prompt in your chat tool:

Create glossaries named "Finance Dictionary" and "Marketing Glossary" with VERIFIED certificate status.

To add subcategories or terms, use this prompt in your chat tool:

Add a subcategory "Revenue Metrics" under the "Finance Glossary".
Add a draft term "Annual Recurring Revenue (ARR)" under the "Finance Glossary".