Improved SQL IntelliSense
Improved SQL IntelliSense is a SQL Server completion provider designed for fast navigation of large databases. It uses case-insensitive Contains matching, so a fragment can match anywhere in an object name, while reusing the active Microsoft mssql connection. No second connection or duplicate credentials are required.
Typing addr may find:
Addresses
CustomerAddresses
BillingAddress
ShippingAddress
[!IMPORTANT]
Disable Microsoft mssql suggestions when using this extension.
Improved SQL IntelliSense replaces the suggestion provider from ms-mssql.mssql. Running both providers can produce duplicate or inconsistent completion results.
Choose Disable globally when prompted on first use, or run Improved SQL IntelliSense: Disable Microsoft SQL Suggestions from the Command Palette.
mssql Quick Info and SQL error checking remain available.
Why Contains?
Matching is contiguous and case-insensitive: addr can occur anywhere in a name. An exact match may be prioritized; otherwise results are grouped by semantic type and sorted alphabetically.
Features
- Case-insensitive Contains completion
- Schemas, tables, views, synonyms, table-valued functions, scalar functions, and stored procedures
- CTEs, temporary tables, and table variables
- Column detail with datatype and
NULL/NOT NULL
- Scalar-function signatures and return types, and stored-procedure signatures
- Alias member completion such as
c.addr
Schema.Object, Database.Schema.Object, and Database..Object navigation
- Same-server cross-database completion and database-wide cross-schema search
sys and INFORMATION_SCHEMA completion
- In-memory metadata caching
- Refresh IntelliSense Metadata after DDL changes
- Show Improved SQL IntelliSense Status plus an optional diagnostic output channel
Usage
Contains
SELECT *
FROM addr
This can suggest any table, view, or table-valued function whose name contains addr, regardless of case.
Alias
SELECT c.addr
FROM dbo.Customers AS c
Schema
SELECT *
FROM reporting.cust
Cross database
Same-server database completion uses databases available through the active mssql connection.
SELECT
c.CustomerId,
o.OrderNumber
FROM CRM.dbo.Customers AS c
JOIN Reporting.sales.Orders AS o
ON o.CustomerId = c.CustomerId
Database-wide editing shortcut
Normal navigation works as follows:
Database. -> schemas
Database.Schema. -> objects from that schema
Database.fragment -> schema matches first, then objects across schemas
If ReportingDb.addr finds CustomerAddresses in schema sales, accepting it inserts:
ReportingDb.sales.CustomerAddresses
The shortcut inserts a valid Database.Schema.Object name; it is not new SQL Server syntax.
System targets work normally, including:
INFORMATION_SCHEMA.TABLES
INFORMATION_SCHEMA.COLUMNS
sys.tables
sys.columns
sys.objects
Use Refresh IntelliSense Metadata after schema changes and Show Improved SQL IntelliSense Status for connection, cache, and completion-provider diagnostics.
Requirements
Privacy
Improved SQL IntelliSense has no telemetry. It does not transmit query text to an external service, store SQL credentials, or open an independent SQL connection. SQL Server catalog access is performed only through the installed Microsoft mssql extension and its active editor connection.
Known limitations
- The defensive tokenizer is not a complete T-SQL compiler. Deeply nested queries or unusual grammar can reduce context accuracy.
- CTE, table-variable, and temporary-table names are recognized, but their columns are not inferred yet.
- Stored-procedure first-result-set discovery is not performed; no result schema is fabricated.
- Metadata refresh after DDL is explicit.
- Cross-database completion is limited to databases on the active SQL Server connection. Linked Servers and four-part names are out of scope.
Development and support
Contributor guidance is maintained in docs/DEVELOPMENT.md and docs/PUBLISHING.md; support guidance is in SUPPORT.md in the source repository.
License
Improved SQL IntelliSense is released under the MIT License. See the root LICENSE file for the full terms.