kusto-mcp
Server Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
No arguments | |||
Instructions
Guidance the server publishes about itself, which clients place ahead of the tool catalog so the model reads it before choosing anything.
This server publishes no instructions, or was last inspected before Glama recorded them.
Capabilities
Features and capabilities supported by this server
Protocol revision2025-11-25
| Capability | Details |
|---|---|
| tools | {} |
| prompts | {} |
Tools
Functions exposed to the LLM to take actions
| Name | Description |
|---|---|
| initialize-connectionB | Creates connection to an ADX cluster |
| show-tablesA | List tables in the current database |
| show-tableB | Show the table schema columns |
| execute-queryA | Runs KQL queries and returns results. No cluster is configured: call initialize-connection before the first query. If you have not inspected the table yet, call show-table first to see its columns. By default, limits results to 20 rows to prevent context overflow. Use the "limit" parameter to specify a different maximum. If results are marked as partial, consider revising your query to use aggregations, filters, or summarizations. Avoid the search operator on large tables; it often times out. Filter a specific table with where/has instead. Optionally set "purpose" to the closest intent. |
| show-functionsB | List functions in the current database |
| show-functionA | Show details of a specific function, including its code and parameters |
| report-issueA | Report a bug or request a feature for kusto-mcp on GitHub. Returns a pre-filled GitHub issue link you open in your browser and submit under your own GitHub account — no GitHub token is used or stored, and no Kusto connection is required. |
Prompts
Interactive templates invoked by user choice
| Name | Description |
|---|---|
| analyze-query-perf | Analyze KQL query performance by measuring scanned extents and totalCpu impact of each query part |
Resources
Contextual data attached and managed by the client
| Name | Description |
|---|---|
No resources | |
TDQS
Scored across 7 tools
The set contains two near-homographic pairs (show-function/show-functions and show-table/show-tables) that differ by a single character, which is a real misselection hazard for an agent. Descriptions do clarify the distinction (detail with code/parameters vs. list), and execute-query vs. show-* is clearly separated, so boundaries are recoverable but not crisp.
All names use kebab-case with a leading verb (initialize-, show-, execute-, report-), which is predictable and readable. The only deviation is the singular/plural suffix convention used to encode detail-vs-list, which is consistent as a rule but collides visually with the base names.
Seven tools is well within the ideal range and each one maps to a distinct step in a Kusto workflow: connect, discover tables/functions, inspect schema, run queries, and report issues. Nothing feels redundant or padded.
The query lifecycle is covered, but the surface lacks database-level discovery (no show-databases) and connection management (no list/switch connection or change-database), even though show-tables/show-functions are scoped to the 'current database'. There is also no way to inspect a function definition's siblings or sample table data other than writing a query, which agents can work around but may stumble on.