Four tools with basic schemas and descriptions, but significant gaps in parameter documentation and output schema completeness. Tool names follow verb_noun convention (list_, query_, get_, return_). Descriptions exist but are terse (10-60 chars, below the 50-200 char baseline for LLM optimization). Only 1 of 4 tools has an outputSchema defined (get_examples). Parameters lack descriptions in most tools: list_queries has 'limit' with minimal context; query_db has 'query' and 'name' with vague guidance ('Valid JSON string'); return_query has 'id' with no format hint. No error handling guidance, no enum constraints, no validation rules. The query_db tool accepts arbitrary JSON strings, creating injection risk without sanitization guidance.
Retrieve example queries for reference
List of recent queries that you performed via the MCP during this session
Execute a database query
Return a query result from recent queries based on an identifier
Missing output schemas for 3 of 4 tools (list_queries, query_db, return_query). LLMs cannot plan downstream calls or extract structured data without knowing response fields.
Parameter descriptions are missing or trivial. 'limit' has no range guidance; 'query' says 'Valid JSON string' without format/example; 'id' has no hint about format or source. LLMs cannot validate inputs without explicit constraints.
query_db accepts arbitrary JSON strings without validation or injection protection guidance. Tool description does not warn about SQL injection risk or sanitization requirements.
Tool descriptions are too short (10-60 chars vs 50-200 char baseline). 'List of recent queries...' and 'Execute a database query' lack context for when/why to use each tool and what they return.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | F | 41 | 2025-06-18+ | v2 |
No error handling guidance. Tools do not document what happens on invalid input, missing resources, or API failures. LLMs have no recovery path.