Loading ToolBench
Serveur MCP pour la traduction de documents. Supporte la traduction de fichiers (PDF, DOCX, TXT, etc.) vers plusieurs langues via Azure Translator, avec intégration OneDrive et Azure Storage.
Server has 4 tools with basic schemas and descriptions, but significant quality gaps. Tool descriptions are in French and lack LLM-optimized structure (no WHAT/WHEN/WHY clarity). Schemas present but incomplete: parameters lack type definitions in several cases, descriptions are minimal or missing. No error handling guidance, no output schemas documented, no parameter constraints (enums, ranges). Tool names are reasonable (verb_noun pattern) but descriptions fail to guide LLM selection. No evidence of idempotency, pagination, or composition patterns. This is a typical community server with functional but poorly-optimized definitions.
Vérifie l'état d'une traduction en cours et retourne le statut actuel.
Récupère le document traduit et l'enregistre sur OneDrive de l'utilisateur.
Liste les fichiers disponibles dans le container doc-to-trad pour traduction.
Lance la traduction d'un document déjà uploadé dans le container doc-to-trad. Le fichier doit être préalablement déposé via Copilot.
Descriptions lack LLM-optimized structure. All 4 tool descriptions are in French, under 100 chars, and omit WHAT/WHEN/WHY guidance. 'Vérifie l'état d'une traduction en cours et retourne le statut actuel' does not explain when to call this vs other tools, what the return format is, or what 'statut' values mean.
No output schemas documented. Tools return data but LLMs cannot plan downstream calls without knowing response structure. E.g., check_translation_status returns 'statut actuel' but no schema shows what fields/types are in the response.
Parameter descriptions are minimal or missing. 'target_language' has description 'Code langue cible (ex: 'fr', 'en', 'es', 'de')' but no enum constraint, no validation rule, no guidance on unsupported languages. 'filter_extension' in list_available_files has no description at all.
Inferred effective spec: <=2025-11-25.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | F | 40 | <=2025-11-25 | v2 |
No error handling guidance. Tools do not document what errors can occur, whether they are retryable, or what the LLM should do next. E.g., translate_document may fail if blob_name is invalid, but no recovery path is provided.
No pagination or result limits documented. list_available_files may return hundreds of files with no limit or cursor guidance. Large unstructured responses waste tokens and degrade LLM reasoning.