Please use the form below to provide your feedback. Because your feedback is valuable to us, the information you submit in this form is recorded in our issue tracking system (JIRA), which is publicly available. You can track the status of your feedback using the ticket number displayed in the dialog once you submit the form.

Managed MCP Server Tools Reference

  • reference
The Couchbase AI Data Plane Managed MCP Server exposes specific tools that you can use with your agents to interact with an operational cluster.

Managed MCP Servers can expose tools across the following categories to an LLM through the MCP protocol:

Cluster Setup and Health

Use cluster setup and health tools to check the status of your Managed MCP Server or its linked operational cluster.

Tool Required API Key Access Description

get_server_configuration_status

Read

Check the configuration status of your Managed MCP Server.

test_cluster_connection

Read

Check your connection to the operational cluster you linked to your Managed MCP Server.

get_cluster_health_and_services

Read

Get information about the health and available Services on your linked operational cluster.

Data Model and Schema Discovery

Use data model and schema discovery tools to explore and modify buckets, scopes, and collections in your linked operational cluster. You can also get the schema for documents in a collection.

If you want to get or modify specific documents, use Document KV Operations tools, instead. If you want to run a query, use Query and Indexing tools.

Tool Required API Key Access Description

get_buckets_in_cluster

Read

Get a list of all buckets on the linked operational cluster.

get_scopes_in_bucket

Read

Get a list of all the scopes in a specified bucket.

get_collections_in_scope

Read

Get a list of all the collections in a specified scope.

get_scopes_and_collections_in_bucket

Read

Get a list of all the scopes and collections in a specified bucket.

get_schema_for_collection

Read

Get an inferred document structure for all the documents in a collection.

create_scope

Read and Write

Create a new scope in a specified bucket.

create_collection

Read and Write

Create a new collection in a specified scope.

delete_scope

Read and Write

Delete the specified scope.

delete_collection

Read and Write

Delete the specified collection.

Document KV Operations

Use document KV operations tools to get and modify specific documents and subdocuments inside your linked operational cluster.

If you want to run a query, use Query and Indexing tools.

Tool Required API Key Access Description

get_document_by_id

Read

Get a document with a specified ID value from a scope and collection.

upsert_document_by_id

Read and Write

Insert a new document or update an existing document with the specified ID value in a scope and collection.

insert_document_by_id

Read and Write

Insert a new document into a scope and collection with a specified ID value. This operation fails if a document already exists with the specified ID value.

replace_document_by_id

Read and Write

Replace an existing document that has the specified ID value. This operation fails if there is no document with the specified ID value.

delete_document_by_id

Read and Write

Delete the document with the specified ID value.

lookup_subdocument

Read

Get parts of a document, such as specific fields, or run existence or array and object counts, without fetching the whole document.

mutate_subdocument

Read and Write

Edit parts of a document, such as specific fields, without rewriting the entire document. You can upsert, insert, replace, remove, run array operations or modify counters.

Query and Indexing

Use query and indexing tools with your linked operational cluster and Managed MCP Server to:

Tool Required API Key Access Description

run_sql_plus_plus_query

Read and Write

Run a SQL++ query using the Query Service on the linked operational cluster.

explain_sql_plus_plus_query

Read and Write

Get information about the execution plan for a SQL++ statement. The explain_sql_plus_plus_query tool includes operators such as scans, joins, and filters. It helps you tune query performance by showing index usage, cost estimates, and data access paths.

list_indexes

Read

Get a list of all indexes in the linked operational cluster, along with their definitions. You can filter indexes by bucket, scope, collection, or index name.

get_index_advisor_recommendations

Read and Write

Get recommendations on the most optimal indexes from the Couchbase Index Advisor for a specific SQL++ query.

create_index

Read and Write

Create a non-vector GSI index for a specified collection. By default, the created index is a deferred index. You cannot use this tool for vector indexes.

build_index

Read and Write

Start building all deferred indexes on a collection.

drop_index

Read and Write

Drop a specified GSI index or vector index from a collection.

Query Performance Analysis

Use query performance analysis tools to identify slow queries, missing indexes, and other optimization opportunities on your linked operational cluster.

Tool Required API Key Access Description

get_longest_running_queries

Read

Get the longest running queries on your linked operational cluster, based on average service time.

get_most_frequent_queries

Read

Get the most frequently executed queries on your linked operational cluster.

get_queries_not_selective

Read

Get any queries from your linked operational cluster that are not selective. The index scans for these queries return many more documents than the final result set.

get_queries_not_using_covering_index

Read

Get any queries from your linked operational cluster that do not use a covering index.

get_queries_using_primary_index

Read

Get any queries from your linked operational cluster that use only a primary index. This can indicate a performance concern with these queries.

get_queries_with_largest_response_sizes

Read

Get the queries from your linked operational cluster that have the largest response sizes.

get_queries_with_large_result_count

Read

Get the queries from your linked operational cluster that have the largest result counts.

See Also