Proactive server health checker — what's about to go wrong?
Single tool with complete schema and clear description, but lacks parameter descriptions, output schema documentation, and error handling guidance. The tool name 'check_health' is generic (not verb_noun pattern like 'run_health_check' or 'get_system_health'). No input parameters means no parameter validation or constraints to document. Output is unstructured text (JSON pretty-printed as string), not a typed response object. No error recovery guidance or actionable error messages. Baseline: 194 chars avg description (this tool: 156 chars, acceptable); 100% of A+ tools document output schemas (this tool: none visible).
Run all server health checks and return results with status (Ok, Warning, Critical, Skipped) for system resources, services, backups, certificates, and open ports.
Output schema not documented. Tool returns text blob of JSON; LLM cannot parse structured fields or plan downstream calls. Baseline: 100% of A+ tools document return types.
Tool name 'check_health' is generic and does not follow verb_noun pattern. 'run_health_check' or 'get_system_health' would be clearer. Baseline: 90% of A+ tools start with action verb.
No error handling guidance. If a check fails (e.g., certificate parsing error, service unreachable), the tool returns unstructured text. LLM cannot determine if error is retryable or user-fixable.
No input parameters means no validation, constraints, or parameter descriptions. While the tool accepts no inputs (correct), the schema should document what the tool checks and what statuses mean (Ok, Warning, Critical, Skipped).
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | D | 55 | 2026-07-28+ | v2 |
Response is unstructured text (JSON stringified), not a typed object. LLM must parse JSON from text field, risking parse errors. Should return structured object with per-check results and status enums.