MCP Directory

Atlan MCP Server

Search, govern, and manage Atlan data catalog assets from AI agents via the Model Context Protocol.

Unverified
stdio (local)
API key
Python
View repo 33 Website

Add to your client

Copy the config for your MCP client and paste it into its config file.

Install / run
uvx atlan-mcp-server

Paste into ~/Library/Application Support/Claude/claude_desktop_config.json

{
  "mcpServers": {
    "atlan-mcp-server": {
      "command": "uvx",
      "args": [
        "atlan-mcp-server"
      ],
      "env": {
        "ATLAN_API_KEY": "<YOUR_API_KEY>",
        "ATLAN_BASE_URL": "https://<YOUR_INSTANCE>.atlan.com",
        "ATLAN_AGENT_ID": "<YOUR_AGENT_ID>"
      }
    }
  }
}

Requires `uv` (the Python package runner). Install it from https://docs.astral.sh/uv/ if `uvx` is not found.

Step-by-step guides: Add to Claude Desktop · Add to Cursor · Add to Windsurf

Before you start

  • An Atlan instance and an Atlan API key (ATLAN_API_KEY)
  • Your Atlan instance base URL (ATLAN_BASE_URL, e.g. https://<your-instance>.atlan.com)
  • Docker (for the Docker install method) OR uv / Python 3.11+ (for the uvx install method)

About Atlan MCP Server

An official MCP server from Atlan that connects AI agents to the Atlan data catalog. Powered by the pyatlan SDK, it provides tools to search and retrieve assets, traverse lineage, update asset descriptions and certifications, manage glossaries (glossaries, categories, terms), create/update/schedule/delete data quality rules, and run SQL queries against table and view assets. It runs locally over stdio (Docker or uvx) for Claude Desktop and Cursor, with SSE and streamable-http transports available for remote/production deployments. A tool-restriction middleware (RESTRICTED_TOOLS) enables role-based access control. Note: this local server is deprecated/maintenance-only; Atlan recommends its hosted MCP endpoint for new integrations.

Tools & capabilities (12)

search_assets

Search for assets based on conditions.

get_assets_by_dsl

Retrieve assets using a DSL query.

traverse_lineage

Retrieve lineage for an asset.

update_assets

Update asset attributes (user description and certificate status).

create_glossaries

Create glossaries.

create_glossary_categories

Create glossary categories.

create_glossary_terms

Create glossary terms.

create_dq_rules

Create data quality rules on Table, View, MaterialisedView, or SnowflakeDynamicTable assets (column-level, table-level, custom SQL).

update_dq_rules

Update existing data quality rules (threshold, priority, conditions, etc.).

schedule_dq_rules

Schedule data quality rule execution for assets using cron expressions.

delete_dq_rules

Delete one or multiple data quality rules by GUID.

query_asset

Execute SQL queries on table/view assets.

When to use it

  • Ask an AI agent to find data assets in your Atlan catalog using natural-language or conditional search.
  • Trace upstream and downstream lineage for a table, column, or other asset.
  • Bulk-update asset descriptions and certification status to improve catalog documentation.
  • Create and organize business glossaries, categories, and terms from natural language.
  • Author, update, schedule (via cron), and delete data quality rules on tables and views.
  • Run SQL queries against table/view assets directly from your MCP client.

Security notes

Requires an Atlan API key (ATLAN_API_KEY) and your instance base URL (ATLAN_BASE_URL) passed as environment variables; store these securely and avoid committing them. The server supports a RESTRICTED_TOOLS environment variable to block specific tools (e.g. for read-only access or to disable raw DSL/SQL queries) — recommended for least-privilege and shared/production deployments. The local server is deprecated and in maintenance-only mode; the vendor recommends the hosted endpoint at mcp.atlan.com/mcp.

Atlan MCP Server FAQ

Do I need Python installed?

It depends on the install method. With Docker (recommended) no Python is needed on the host — the container includes everything. With uv, a Python runtime is required, but uv will automatically download and manage Python 3.11+ for you. The server is implemented in Python and requires Python 3.11+ to run.

Which MCP clients are supported?

The README documents mcpServers config blocks for Claude Desktop and Cursor over stdio. SSE and streamable-http transports are also supported for remote/web deployments.

Can I restrict which tools are exposed?

Yes. Set the RESTRICTED_TOOLS environment variable to a comma-separated list of tool names to hide and block them — useful for read-only access or disabling raw DSL/SQL queries. If unset, all tools are available by default.

Is this server still maintained?

The local server in this repository is marked deprecated and in maintenance-only mode (no new features, support not guaranteed). Atlan recommends its hosted MCP endpoint at mcp.atlan.com/mcp for new integrations.

#atlan#data-catalog#data-governance#lineage#data-quality#metadata#ai-agents

Alternatives to Atlan MCP Server

Compare all alternatives →
Memory (Knowledge Graph)
AI, Data & Knowledge
74k

Official MCP server providing persistent, file-backed knowledge-graph memory across sessions.

Verified
stdio (local)
No auth
TypeScript
9 tools
Updated 5 months agoRepo
Sequential Thinking
AI, Data & Knowledge
62k

Structured step-by-step reasoning tool for breaking problems into revisable thought sequences.

Verified
stdio (local)
No auth
TypeScript
1 tool
Updated 6 months agoRepo
AWS Knowledge MCP
AI, Data & Knowledge
7.0k

Fully managed remote server for AWS docs, blogs, What's-New and Well-Architected guidance — no key.

Verified
HTTP (remote)
No auth
Hosted
6 tools
Updated 5 months agoRepo

Compare Atlan MCP Server with:

vs Memory (Knowledge Graph)vs Sequential Thinkingvs AWS Knowledge MCPvs MCP Server Chart