This STDIO-only demo server exhibits significant definition quality gaps. While tool schemas are present and correctly formatted, descriptions are minimal (averaging ~70 chars, below the 194-char baseline), parameter descriptions are entirely absent, and error handling provides no recovery guidance. All four tools lack input parameter annotations. The server demonstrates basic schema structure but falls well short of production-grade tool definition quality. The implementation uses hardcoded French text, limiting internationalization. No tool annotations (readOnlyHint/destructiveHint) are present despite clear distinctions between read and write operations.
Contrôle une application (restart uniquement)
Affiche les informations de votre compte Clever Cloud
Retourne le nombre d'applications sur Clever Cloud
Retourne la date et l'heure actuelles
Parameter descriptions entirely missing. All four tools declare input parameters (appId, action in control_clever_app) but provide zero descriptions of what each parameter means, expected format, or valid values. LLMs cannot infer parameter intent from names alone.
No output schema documented. While tools return strings (via handler functions), there is no formal specification of return type, structure, or fields. LLMs need documented output schemas to plan downstream tool calls.
Error messages provide no recovery guidance. Handlers throw generic errors ('Failed to fetch Clever Cloud applications', 'Not connected to Clever Cloud') without actionable next steps. Per pattern:recovery-guide, errors must tell the LLM what to do next.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-22 | F | 45 | 2026-07-28+ | v2 |
| 2026-03-09 | F | 41 | - | v1 |
Tool annotations missing. control_clever_app is a destructive operation (restart action) but is not marked with destructiveHint. get_* tools are read-only but lack readOnlyHint annotations. These hints guide agent safety logic.
Tool descriptions are under 100 characters and lack situational context. 'Retourne la date et l'heure actuelles' (52 chars) and 'Affiche les informations de votre compte Clever Cloud' (54 chars) do not explain WHEN to call the tool or differentiate it from similar tools. Baseline is 194 chars for descriptions.
French-language descriptions and hardcoded text limit internationalization and multi-language agent support. Tool descriptions should be in English for maximum LLM comprehension and ecosystem interoperability.