# Atlan MCP for AI Agents MCP

> Atlan MCP connects your AI agent directly to the Atlan Data Catalog, turning metadata discovery and governance into a conversation. Your agent can search thousands of data assets, verify business terms from organizational glossaries, and check data sensitivity classifications (like PII) instantly—all without you having to click through dozens of dashboards.

## Overview
- **Category:** brain-trust
- **Price:** Free
- **Tags:** metadata-management, data-governance, data-catalog, business-glossary, data-stewardship, asset-discovery

## Description

Atlan MCP lets your AI client become a master data steward for your organization. Instead of manually logging into different systems just to figure out where a piece of data lives or what it means, your agent handles the heavy lifting. It connects conversational governance and instant cataloging directly into your workflow. By subscribing through Vinkius, you give your AI agent deep context about every table, column, and business definition across your entire data estate.

Your agent can locate assets using natural language search, retrieve verified terms from internal glossaries, or list active security tags like Confidentiality or GDPR compliance. It's governance on demand. You’ll stop wasting time hunting down the right dataset and start working with trusted information.

## Tools

### search_assets
Searches the entire Atlan Data Catalog for specific data assets by name or keyword.

### list_glossaries
Retrieves a list of all business glossaries defined within your organization's metadata.

### list_classifications
Lists every active data classification tag, such as Confidential or PII, configured in Atlan.

### list_purposes
Shows all defined sharing purposes and policies for your data assets within the platform.

### list_personas
Retrieves a list of configured access control profiles, known as Personas, used in Atlan.

### list_users
Lists all individual users registered within the Atlan workspace.

## Prompt Examples

**Prompt:** 
```
What are the most sensitive types of data we have?
```

**Response:** 
```
**Data Sensitivity Report**

We currently have 7 active classification tags applied across our assets. These dictate how the data must be handled:

*   🛡️ **PII:** Personal Identifiable Information (High Risk)
*   🔒 **Confidential_Tier_1:** Internal use only, restricted access.
*   🌐 **Public_Domain:** Safe for external reporting.

If you want to check which tables contain PII specifically, let me know.
```

**Prompt:** 
```
List all the official business metrics we use?
```

**Response:** 
```
**Verified Business Glossaries**

Our organization maintains 4 primary active glossaries. These define our core KPIs and concepts:

*   `Core Finance KPIs`: Defines 'EBITDA' (Earnings Before Interest, Taxes, Depreciation, and Amortization).
*   `Marketing Attribution Rules`: Provides the official definition for 'First Touch Conversion.'
*   `Engineering Data Contracts`: Specifies how data fields should be named.
*   `Regulatory Policies`: Tracks definitions related to regional compliance mandates.
```

**Prompt:** 
```
Who has access to the executive dashboard?
```

**Response:** 
```
**Access Profile Summary**

The following Personas are currently configured for high-level visibility:

*   `Executive Tier`: Read-only access across all core dashboards.
*   `Data Steward`: Full metadata read/write rights, limited data access.
*   `Finance Analyst`: Limited to Finance schemas and reporting tools.

You can see the full list of users under the `list_users` tool if needed.
```

## Capabilities

### Search Across Data Assets
Find any data asset, whether it's a table, dashboard, or column, by querying its name or topic directly.

### Verify Business Definitions
Retrieve comprehensive organizational glossaries that hold verified business terms and acronyms used company-wide.

### Check Data Policies and Tags
List all configured data classifications (like PII) or review the sharing policies applied across your datasets.

### Inspect User Roles and Access Rights
View registered users in the workspace and check the specific access control profiles assigned to different business roles.

## Use Cases

### A new project needs a reliable KPI definition
Instead of asking five different domain experts, the analyst asks their agent to check for verified business terms. The agent uses `list_glossaries` and immediately returns the official definition and owner for 'Monthly Recurring Revenue,' solving weeks of debate in minutes.

### Preparing for a compliance audit
The security officer asks their agent to list all data classifications. The agent uses `list_classifications` and confirms that every asset containing customer names is tagged as 'PII,' generating an immediate, auditable report.

### Finding the source of truth for a dashboard
A data scientist needs to know which tables feed the executive summary dashboard. They ask the agent to search assets, and it uses `search_assets` to trace the lineage back to the original production schema, identifying the authoritative source.

### Onboarding a new team member
The manager needs to know who has access to sensitive HR data. They ask the agent to list all users and then check the relevant Personas using `list_personas`, giving an immediate overview of role-based permissions.

## Benefits

- Find the right data without clicking through menus. Use `search_assets` to query your entire connected data ecosystem using plain text, instantly revealing lineage and trusted datasets.
- Stop arguing over definitions. The agent uses `list_glossaries` to pull verified business terms and acronyms from organizational glossaries, ensuring everyone speaks the same technical language.
- Stay compliant by design. You can call `list_classifications` to see every active data tag (like PII or Confidential) applied across your assets, guaranteeing proper handling.
- Understand who sees what. By running `list_personas`, you get a clear map of configured access control profiles, allowing security teams to audit permissions quickly.
- Improve trust in the data by listing all defined sharing purposes (`list_purposes`), showing exactly why and how your data can be used.

## How It Works

The bottom line is that your AI client talks to the metadata layer directly, giving it full visibility into what data exists, who owns it, and how sensitive it is.

1. Subscribe to this secure MCP connection on Vinkius.
2. Provide your Atlan API URL and corresponding API Key credentials.
3. Start talking to the catalog via any compatible AI client, letting your agent do the searching and governance work.

## Frequently Asked Questions

**How does Atlan MCP help me find data assets I don't know exist?**
It acts like a universal search engine for your entire company data catalog. Instead of guessing names, you ask your agent a question about the business problem, and it uses `search_assets` to surface every potential dataset that matches the context.

**I need to know if my data is compliant; can Atlan MCP check classifications?**
Yes. Your agent runs classification checks by listing all tags (like PII or GDPR). This gives you a clear, auditable view of your compliance posture across the entire estate.

**What is the difference between Atlan MCP and just using an internal wiki?**
A wiki only stores text; this MCP connects to live metadata. It doesn't just tell you *what* a term means, it tells you *where* that term is used in actual production tables and dashboards.

**Can Atlan MCP help me understand the company's official metrics?**
Absolutely. It accesses your business glossaries via `list_glossaries`. This means you get the single, verified definition for any KPI—like 'Net Revenue' or 'Churn Rate'—so everyone is on the same page.

**I need to know which teams can use a specific dataset. Can Atlan MCP check access rights?**
Your agent lists configured Personas and users, showing you exactly what roles exist and what level of access they have been granted across your catalog.