Read-only SQL Server access over MCP with validated queries, schema tools, ER diagrams, and DBA diagnostics.
Copy the AI prompt to install this server into Claude Code, Cursor, or another agent — or use 1-click editor setup below.
One-click editor setup isn’t available for this listing yet — we don’t have a confirmed install command, and we’d rather show nothing than point your editor at the wrong package or host. Follow the project’s own setup instructions, linked above.
Inspect callable tools, capabilities, and parameters exposed to AI agents by SqlAugur.
list_serversLists available SQL Server instances configured in `appsettings.json`.
list_databasesLists all databases on a named server with names, IDs, states, and creation dates.
read_dataExecutes a read-only SQL SELECT query. Only `SELECT` and `WITH` (CTE) queries are allowed. Results returned as JSON with a configurable row limit.
get_query_planReturns the estimated or actual XML execution plan for a SELECT query.
get_schema_overviewConcise Markdown schema overview: tables, columns, PKs, FKs, unique/check constraints, defaults. Supports `compact` mode, schema and table filtering.
describe_tableComprehensive table metadata in Markdown: columns, data types, nullability, defaults, identity, computed expressions, indexes, FKs, constraints.
The mbentham/SqlAugur MCP server connects an MCP client to one or more named SQL Server instances. Connections are defined in appsettings.json, and list_servers exposes the configured names for later tool calls. Database discovery includes database names, IDs, states, and creation dates.
Its query interface is intentionally read-only. read_data accepts only SELECT statements and common table expressions beginning with WITH, then returns JSON subject to a configurable row limit. The server also provides estimated or actual XML query plans for SELECT statements. Query validation uses Microsoft's T-SQL parser to inspect the full syntax tree rather than relying on keyword matching.
The MCP client communicates with the server over standard input and output. SqlAugur validates incoming queries before sending permitted work to SQL Server. Diagnostic stored procedure calls have parameter restrictions intended to prevent write operations. Throughput and concurrency controls are included to limit repeated or simultaneous agent requests.
Core database tools support schema inspection and documentation. Agents can retrieve a compact or detailed schema view, describe a table's columns and constraints, inspect views, procedures, functions, and triggers, read object definitions and extended properties, and examine upstream or downstream dependencies. Diagram tools write PlantUML or Mermaid ER definitions to .puml or .mmd files and support schema, table, compactness, and table-count options.
The documented installation options are the NuGet global tool, a Docker or Podman container, and a source build. The global tool requires the .NET 10 runtime and is installed with dotnet tool install -g SqlAugur; the executable used in MCP client configuration is sqlaugur. Source builds require the .NET 10 SDK.
For local installation, place appsettings.json under ~/.config/sqlaugur/ on Linux or macOS, or under %APPDATA%\\sqlaugur on Windows. Container deployments can mount the file at /app/appsettings.json or provide nested SqlAugur__Servers__...__ConnectionString environment variables. Each configured server needs a named connection string. SQL Server connectivity and authentication depend on those connection strings.
The initial toolset includes server and database listing, read-only data access, query plans, and schema overview. Additional tools can be discovered through list_toolsets and inspected with get_toolset_tools before being enabled.
Available diagnostic coverage includes overall health checks, DMV-based performance sampling, plan-cache and Query Store analysis, index review, active-session and blocking inspection, deadlock and system-health parsing, error-log searches, and query reproduction script generation. The listed integrations include SQL Server First Responder Kit, DarlingData, and sp_WhoIsActive; availability depends on the corresponding database-side setup.
Only SELECT and CTE queries are accepted by read_data; this interface is not intended for INSERT, UPDATE, DELETE, DDL, or other mutations. PlantUML and Mermaid generation enforce a maximum of 200 tables. Diagnostic outputs may omit verbose columns and truncate long strings by default, with documented options such as verbose and includeQueryPlans available where supported.
The mbentham/SqlAugur MCP server requires a reachable SQL Server and correctly configured connection. Its cross-server plan comparison tool requires the demon_hunters branch of First Responder Kit until the referenced change is merged. The project is licensed under MIT.
Factual signals from GitHub, npm, and our automated checks — not a rating.
No reviews yet — be the first to share how this listing worked for you.
Showcase your server listing on GitHub or your project documentation. Embed this dynamic SVG badge to highlight official listing status and live engagement.
[](https://allmcps.com/mcp/mbentham-sqlaugur)<a href="https://allmcps.com/mcp/mbentham-sqlaugur"><img src="https://allmcps.com/api/badge/mbentham-sqlaugur?style=directory" alt="SqlAugur on AllMCPs" /></a>