Couchbase (Vector & NoSQL) MCP. Query structured data and search vectors from natural language.
Gemini Works with every AI agent you already use
…and any MCP-compatible client
Just plug in your AI agents and start using Vinkius.
Couchbase (Vector & NoSQL) MCP Server connects your AI agent directly to your Couchbase cluster. It lets you run N1QL queries against structured JSON data, perform KNN vector similarity searches, and fetch specific documents by key.
You can also list schemas, audit indexes, and run full-text searches—all from natural conversation.
What your AI agents can do
Execute n1ql query
Runs structured JSON queries using N1QL syntax against your database.
Fts search
Performs keyword-based text extraction matching structured content trees.
Get document
Fetches the complete data map of a document using its unique key.
Runs structured SQL queries against your JSON data to retrieve specific fields and records.
Compares provided vectors to your indexed data to find semantically similar content.
Fetches an entire document's properties using its unique ID key.
Searches documents using keywords across structural indexes.
Lists and navigates your entire data structure, including buckets, scopes, and collections.
Lists all search indexes and verifies the underlying vector and cluster definitions.
Ask AI about this MCP
Supported MCP Clients
Gemini Waiting for input…
Couchbase (Vector & NoSQL) MCP Server: 7 Tools
These tools let your AI client interact with Couchbase, enabling structured querying, semantic search, and data schema auditing.
019d757dexecute n1ql query
Runs structured JSON queries using N1QL syntax against your database.
019d757dfts search
Performs keyword-based text extraction matching structured content trees.
019d757dget document
Fetches the complete data map of a document using its unique key.
019d757dlist buckets
Lists all top-level data containers (buckets) in the database.
019d757dlist indexes
Enumerates all active search indexes attached to the cluster.
019d757dlist scopes
Retrieves the specific organizational boundaries (scopes and collections) within a bucket.
019d757dvector search
Calculates and lists the nearest neighbors based on vector embeddings.
Choose How to Get Started
Build a custom MCP for your own tools, or connect a ready-made integration from our catalog.
Build Your Own
Turn any API into an MCP. Import a spec, define Agent Skills, or deploy with MCPFusion.
- Import from OpenAPI, Swagger, or YAML specs
- Create Agent Skills with progressive disclosure
- Deploy to edge with MCPFusion framework
- Built in DLP, auth, and compliance on every call
- Real time usage dashboard and cost metering
- Publish to catalog or keep private
Make Your AI Do More
Start with Couchbase (Vector & NoSQL), then connect any of our 4,700+ other servers whenever your AI needs more. One click, no limits.
- Use this MCP plus 4,700+ others, all in one place
- Add new capabilities to your AI anytime you want
- Every connection is secured and compliant automatically
- Track usage and costs across all your servers
- Works with Claude, ChatGPT, Cursor, and more
- New servers added to the catalog every week
What you can do with this MCP connector
Your AI client connects straight to your Couchbase cluster. You'll get full control over your NoSQL and vector data using natural conversation.
execute_n1ql_query runs structured JSON queries using N1QL syntax against your database, letting you pull specific fields and records. vector_search calculates and lists the nearest neighbors based on vector embeddings, finding content that's semantically similar. get_document fetches the complete data map of a document using its unique key. list_buckets lists all top-level data containers (buckets) in the database. list_indexes enumerates all active search indexes attached to the cluster. list_scopes retrieves the specific organizational boundaries (scopes and collections) within a bucket. fts_search performs keyword-based text extraction matching structured content trees.
How Couchbase (Vector & NoSQL) MCP Works
- 1 Subscribe to the server and provide your Couchbase URL, Database Username, and Database Password.
- 2 Tell your AI client what data you need—for example, 'Find all flights under $100' or 'Show documents similar to X'.
- 3 Your agent translates that request into the appropriate tool call (e.g.,
execute_n1ql_queryorvector_search) and returns the resulting data set.
The bottom line is that your AI client talks directly to Couchbase, and the server runs the complex database logic for you.
Who Is Couchbase (Vector & NoSQL) MCP For?
The data architect who can't afford to spend weeks manually checking schemas. The developer who needs to test vector similarity for a new RAG app without setting up a local database. The DBA who needs to audit indexes quickly before a major migration.
Verifies collection organization and audits NoSQL data structures across different environments.
Tests and debugs semantic matching and vector similarity searches for Retrieval-Augmented Generation (RAG) apps.
Executes N1QL queries to verify data consistency and monitors search index health.
What Changes When You Connect
- Deep Semantic Search: Don't just look for keywords. Use
vector_searchto find documents that mean the same thing, even if they use different words. It handles semantic similarity. - Structured Data Access: Need a specific field from a complex JSON object? Use
execute_n1ql_queryto push targeted SQL queries and get clean results, skipping the manual data extraction. - Schema Visibility: Don't waste time guessing where your data is. Run
list_bucketsandlist_scopesto map the entire data hierarchy instantly. - Document Integrity Checks: Use
get_documentto pull the full, exact data map of any document by its key. This is crucial for debugging or validating data state. - Full Text Indexing: Use
fts_searchwhen you need robust, structural text extraction based on common search queries, separate from vector matching. - Audit Readiness: Before a migration or feature launch, run
list_indexesto verify that all necessary search indexes and vector definitions are active and correct.
Real-World Use Cases
Debugging a Bad Search Result
A developer gets a search result, but it's missing context. They first run get_document using the document key to pull the full internal property map. They then use list_scopes to confirm the document belongs to the right collection. This quickly isolates whether the issue is the query or the underlying data structure.
Finding Related Concepts in a Corpus
A researcher has a new concept and needs to see similar articles. Instead of keyword searching, they convert the concept into a vector and run vector_search. This returns the top 5 most semantically related documents, regardless of whether the original article used the exact terminology.
Building a Limited Scope View
A product manager needs a report showing only active users' names and last login dates, limited to the 'US' region. They use execute_n1ql_query to write a precise N1QL query that filters the complex JSON structure, avoiding a full table scan.
Mapping a New Data Source
A data architect inherits a new database. They start by running list_buckets and list_scopes to map the entire structure. Then, they use list_indexes to see what search indexes already exist, establishing a baseline audit.
The Tradeoffs
Treating everything like a simple keyword search
Running only fts_search assuming it will capture semantic meaning. This fails if the user meant 'economic stability' but the document only uses 'financial resilience.'
→
For semantic matching, always use vector_search and provide the appropriate embedding. Use fts_search only when you need strict keyword matching against indexed content.
Assuming the data is flat
Trying to query nested JSON fields without knowing the schema, leading to N1QL errors or incomplete results.
→
Before running execute_n1ql_query, run list_scopes and list_indexes to confirm the correct data hierarchy. Check the schema first.
Over-querying the whole cluster
Running a full scan (SELECT * FROM bucket) just to find a few pieces of data. This is slow and hits resource limits.
→
Always start by running list_indexes to identify the correct search index, then use fts_search or vector_search to narrow the scope before querying the data.
When It Fits, When It Doesn't
Use this server if your data is complex, structured JSON and you need to query it using both exact fields (N1QL) and semantic meaning (vector embeddings). Don't use it if your data is in a simple, relational SQL database—you need a standard SQL connector. If you only need to search by keyword and don't care about deep meaning, a basic search engine might suffice, but Couchbase offers the added benefit of running the search and retrieving the underlying JSON data in one flow. The key is that you're dealing with heterogeneous data: you need the structural checks (list_scopes), the full-text search (fts_search), the exact query (execute_n1ql_query), and the deep context (vector_search).
Independent Platform Disclaimer: Vinkius is an independent platform and is not affiliated with, endorsed by, sponsored by, verified by, or otherwise authorized by Couchbase. All third-party trademarks, logos, and brand names are the property of their respective owners. Their use on this website is strictly for informational purposes to identify service compatibility and interoperability.
VINKIUS INFRASTRUCTURE
Cloud Hosted
Managed infra
V8 Isolated
Sandboxed per request
Zero-Trust Proxy
No stored credentials
DLP Enforced
Policy on every call
GDPR Compliant
EU data residency
Token Compression
~60% cost reduction
Works with Claude, ChatGPT, Cursor, and more
The Model Context Protocol standardizes how applications expose capabilities to LLMs. Instead of operating in isolation, your AI gains direct access to external platforms, live data, and real-world actions through secure, standardized connections.
This server provides 7 capabilities that interface natively with Claude, ChatGPT, Cursor, and any MCP client. No middleware. No custom integration required.
Available Capabilities
Manually checking data structure is a time sink.
Right now, figuring out where a piece of data lives—whether it's in the 'user' bucket, under the 'production' scope, or the 'finance' collection—is a tedious click-through process. You jump through dashboards, check schema documentation, and copy-paste identifiers just to confirm the path.
With this MCP server, you ask your agent to map the data structure. It runs `list_buckets` and `list_scopes` and gives you a clean, immediate report of the entire data map. You know exactly where to look, instantly.
Couchbase (Vector & NoSQL) MCP Server: Full Control.
You no longer have to switch between dedicated search tools, SQL consoles, and vector databases. You tell your agent, 'Find all high-value products mentioned in documents related to Q4 earnings.' The agent orchestrates the necessary calls—it runs `vector_search` first to find related documents, then uses `execute_n1ql_query` to pull the product data from those documents.
The difference is orchestration. You get a single, conversational endpoint that handles multiple, specialized database tasks—from schema mapping to semantic retrieval—in one go.
Common Questions About Couchbase (Vector & NoSQL) MCP
How do I use the `vector_search` tool with Couchbase (Vector & NoSQL) MCP Server? +
You provide the vector embedding and a query context. The tool finds the closest matches using indexed vector data. This is for semantic similarity, not keyword matching.
What's the difference between `execute_n1ql_query` and `fts_search`? +
execute_n1ql_query is for structured querying using SQL on JSON fields. fts_search is for keyword-based text extraction across structural indexes.
Can I check the schema using the `list_scopes` tool? +
Yes, list_scopes retrieves the explicit organizational limits (scopes and collections) within a bucket. This tells you the finer detail of the data organization.
What if I need to find a document by its ID? +
Use the get_document tool. You supply the unique document key, and it fetches the entire internal property map for that specific record.
How do I check what data is available using the `list_buckets` tool? +
The list_buckets tool shows the top-level containers in your cluster. It returns a list of all available buckets, which defines your primary data routing spaces. You'll need to select a bucket before running any queries.
What is the purpose of the `list_indexes` tool? +
The list_indexes tool enumerates all active search indexes attached to your database. This lets you verify which structured rules are exporting data, ensuring your subsequent searches or queries target the correct index.
Can I use the `get_document` tool to audit my data structure? +
Yes, get_document retrieves specific documents using a unique key. This is the best way to check a document's internal properties and verify its exact data map within a collection.
What does the `execute_n1ql_query` tool require for proper setup? +
You must provide your Couchbase URL, username, and password in the server settings. The tool then executes N1QL queries against the JSON payloads, requiring accurate scope and collection names in your query syntax.
Can my agent perform K-Nearest Neighbor (KNN) vector searches in Couchbase? +
Yes. Provide the search index name, the vector embedding array, and the number of results (k). The agent uses Couchbase's native vector capabilities to locate the most semantically similar documents in your cluster.
How do I execute a N1QL query through the agent? +
Use the 'execute_n1ql_query' tool and provide your SQL-like statement. The agent will fetch the structural JSON blocks directly from Couchbase, allowing you to perform complex data retrieval using familiar SQL syntax.
Can I search documents using full-text query logic? +
Absolutely. The 'fts_search' tool leverages Couchbase's Full-Text Search (FTS) engine. Provide an index name and a boolean query string to perform structural text-based extraction across your document trees.
Use it with your favorite AI tools
Connect this server to Cursor, Claude, VS Code, and more.
More in this category
Alchemy
Manage blockchain data — audit wallets, NFTs, and transactions via AI.
TiDB Cloud (Serverless Distributed SQL)
Manage TiDB Cloud infrastructure — list projects, inspect clusters, and monitor serverless instances directly from your AI agent.
Deta Space (Serverless Personal Cloud API)
Manage your personal cloud via Deta Space — store data in Deta Base and files in Deta Drive directly from any AI agent.
You might also like
Fusebill (Enterprise Subscription Billing API)
Automate enterprise subscription billing via Fusebill — manage customers, track subscriptions, and handle billing lifecycles directly from any AI agent.
WellnessLiving
Manage schedules, clients, staff, memberships, locations, and reports for your WellnessLiving-powered studio through natural conversation.
Porkbun
Manage domains and DNS records via Porkbun — list domains, update DNS, check pricing, and retrieve SSL certificates directly from your AI agent.