Hologres MCP Server with STDIO and SSE transport modes support
Server has 4 tools with basic schemas and descriptions, but significant quality gaps. Tool descriptions are minimal (10-50 chars, well below the 194-char baseline). Parameter descriptions are missing entirely, the schema shows property names but no descriptions for 'query', 'schema', or 'table'. No output schemas documented. No error handling guidance. execute_sql and analyze_table are WRITE operations but descriptions do not state this, violating the command-tool pattern. No input validation, no recovery guidance, no pagination support despite SQL tools potentially returning large result sets.
Collect statistics for a table in Hologres
Execute SQL query in Hologres database
Get execution plan for a SQL query
Get query plan for a SQL query
Parameter descriptions completely missing. Schema defines 'query', 'schema', 'table' properties but provides no descriptions. LLMs cannot infer parameter meaning from names alone.
Tool descriptions are too short (10-50 chars vs 194-char baseline). 'Execute SQL query in Hologres database' lacks context on when to use it, what it returns, and whether it modifies state. Descriptions must answer: What? When? What does it return?
WRITE operations (execute_sql, analyze_table) do not declare their destructive nature in descriptions. Agents need to know which calls are safe to retry and which have irreversible consequences.
Inferred effective spec: <=2025-11-25.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | F | 39 | <=2025-11-25 | v2 |
No output schemas documented. LLMs cannot plan downstream tool calls or extract the right data without knowing what fields to expect from execute_sql or analyze_table results.
No error handling guidance. Raw errors or stack traces give agents nothing to act on. Missing recovery hints like 'If query fails, check syntax with get_query_plan()' or 'If table not found, call list_tables() first.'