MCP server for executing Redash queries and retrieving results
Single tool with minimal schema documentation. Tool name 'execute_query' lacks action clarity (execute is vague). Input schema present but parameter descriptions are missing entirely, 'params' has no description of what parameters the Redash query accepts. Output schema undocumented; response structure inferred from code only. Tool description is generic (10 words, 45 chars) and does not explain when to use it, what it returns, or prerequisites. No error guidance for LLM recovery. Baseline: A+ tools average 194 chars description + 100% param descriptions; this tool has neither.
Execute Redash query and get results
Parameter descriptions missing. 'params' parameter has no description explaining what Redash query parameters it accepts, format, or constraints. LLM cannot infer valid input.
Output schema undocumented. Response structure (JSON with 'query_result' → 'data' nesting) is not declared in tool definition. LLM cannot plan downstream operations or extract fields reliably.
Tool description too generic (45 chars). Does not explain when to call execute_query vs alternatives, what Redash is, or what data structure is returned. Baseline: 194 chars average for A+ tools.
No error guidance for LLM recovery. Errors like 'Query execution failed' or 'Job status check failed' do not tell the agent what to do next (retry? ask user? check query ID?). Baseline: A+ tools include recovery hints.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | F | 30 | 2026-07-28+ | v2 |
query_id parameter lacks description. No explanation of what a valid query_id is, how to obtain it, or whether it is a Redash internal ID or user-facing name.