Atlan

Create data quality rules in Atlan. Always uses propose mode — STOP after proposing and wait for user approval before executing.

Create one or more glossaries in Atlan. Always uses propose mode — STOP after proposing and wait for user approval before executing.

Create one or more glossary categories in Atlan. Always uses propose mode — STOP after proposing and wait for user approval before executing.

Create one or more glossary terms in Atlan. Always uses propose mode — STOP after proposing and wait for user approval before executing.

Permanently delete one or more custom metadata sets and all their attribute values from assets. Irreversible. Always uses propose mode — STOP after proposing and wait for user approval before executing.

Delete data quality rules by their GUIDs. Always uses propose mode — STOP after proposing and wait for user approval before executing.

Get workspace groups and their members. Use include_members=True to list users in a group.

Add or remove announcements (information, warning, issue) on assets. Always uses propose mode — STOP after proposing and wait for user approval before executing.

Manage asset lifecycle: archive, restore, or permanently purge assets. WARNING: PURGE cannot be undone. Always uses propose mode — STOP after proposing and wait for user approval before executing.

Execute SQL queries against Atlan connections to preview data.

Remove an Atlan tag from one or more assets. Pass a dict for single asset, list for batch. Always uses propose mode — STOP after proposing and wait for user approval before executing.

Remove (archive) attributes from an existing custom metadata set. Archived attributes are soft-deleted and their values cleared from all assets. Always uses propose mode — STOP after proposing and wait for user approval before executing.

Remove a custom metadata set from an asset. Always uses propose mode — STOP after proposing and wait for user approval before executing.

Use this to discover metadata entities by name or description: users, groups, classifications, custom metadata (business_metadata) sets, glossaries, domains, and data products. Use before write operations to confirm exact names and GUIDs. NOT needed to read CM values from an asset — pass the CM set name in include_attributes on get_asset_tool instead.

Schedule data quality rule execution on assets. Always uses propose mode — STOP after proposing and wait for user approval before executing.

BACKUP search tool — only use when the primary search tools (start_deep_query_tool or semantic_search_tool) cannot express what you need. Provides structured asset search with exact filters, conditions, aggregations, sorting, and pagination. Use this ONLY for precise attribute filters (e.g. certificateStatus, connectorName), term_guids, tags, domain_guids, count_only, or aggregations that the primary tools don't support. LINKED/UNLINKED TERMS — SINGLE CALL: To find linked or unlinked glossary terms, make ONE call: search_assets(asset_type="AtlasGlossaryTerm", glossary_qualified_name="<qn>", aggregations={"linked_terms": {"field": "__meanings"}}) The server automatically fetches ALL terms (no limit cap), checks which have linked assets, and returns: - linked_terms[]: each with name, guid, qualifiedName, asset_count, and assets[] (up to 5 linked assets with full details) - unlinked_terms[]: each with name, guid, qualifiedName - linked_count, unlinked_count, total_terms No multiple calls or manual diffing needed. Use this for "show unlinked terms", "which terms have linked assets", "orphan terms", etc. GLOSSARY TERM WORKFLOWS: - Filter terms by category (including sub-categories): use category_qualified_name="<categoryQualifiedName>". The server automatically traverses the full hierarchy and returns terms from all descendant categories. NOT for users/groups — use resolve_metadata_tool for those.

Search Atlan's customer-facing documentation and return an LLM-generated answer with source citations. Use for how-to questions about Atlan features — not for searching data assets (use semantic_search_tool for that).

PREFERRED search tool — use this FIRST for any discovery or lookup query. Handles natural language, fuzzy matching, typos, abbreviations, multi-word names, and glossary term lookups. Covers all asset types: tables, columns, views, schemas, dashboards, glossary terms, categories, domains, data products, and more. NOT for users or groups — use resolve_metadata_tool for those. Only fall back to search_assets_tool if you need exact attribute filtering, aggregations, or structured conditions that semantic search cannot express. This tool accepts ONLY: user_query (required), limit, offset, include_readme. It does NOT accept asset_type, conditions, search_query, or any other parameters.

Traverse upstream or downstream lineage from an asset. Returns a widget-ready graph with data assets connected by lineage edges. Process connector nodes are bridged in the graph (A -> B instead of A -> Process -> B); their entity data is exposed in the ``process_map`` field and each relation carries ``via_process_guids`` identifying which ETL/transformation processes produced the connection. The lineage widget supports progressive loading — leaf nodes show an expand button that fetches deeper lineage on demand. Keep size small (5-15) for responsive results; do NOT request all lineage at once.

Update attributes on one or more assets. Each item specifies the asset identity + fields to change. Always uses propose mode — STOP after proposing and wait for user approval before executing.

Update custom metadata on one or more assets. Set replace=True for full replacement, False (default) for partial update. Always uses propose mode — STOP after proposing and wait for user approval before executing.

Update existing data quality rules. Always uses propose mode — STOP after proposing and wait for user approval before executing.