Server has 8 tools with explicit schemas and descriptions, but quality is significantly below production baseline. All descriptions are in Arabic (non-English), which limits LLM comprehension in English-first environments. Schemas are present but lack critical constraints (enums, ranges, patterns). Parameter descriptions are minimal (avg ~30 chars vs baseline 72). No output schemas documented. Error handling is present but not recovery-focused. Tool naming is reasonable (verb_noun pattern) but some names conflate multiple concerns (delete_file_or_folder, confirm_or_cancel_deletion). Security controls exist (path validation, forbidden roots) but are not declared in tool metadata. No tool annotations (readOnlyHint, destructiveHint, idempotentHint) despite high-risk operations.

• Translate all tool and parameter descriptions to English. Use clear, concise language optimized for LLM comprehension (50-200 chars per description, 50-100 chars per parameter).
• Add explicit output schemas for every tool. Document the structure of success responses (e.g., write_file returns {type: 'text', text: string, file_path: string, size_bytes: number}).
• Add tool annotations: mark write_file, edit_file_text, create_directory, move_or_rename_file, delete_file_or_folder as destructiveHint=true; mark read_file_preview, list_directory_contents as readOnlyHint=true; mark confirm_or_cancel_deletion as idempotentHint=true.
• Split delete_file_or_folder + confirm_or_cancel_deletion into a clearer workflow: request_file_deletion (returns request_id) → confirm_deletion (accepts request_id + action). Document the two-step pattern in descriptions.
• Add parameter constraints: file_path should document allowed/forbidden patterns; max_lines should have min/max bounds (1-10000); extension should accept empty string or valid file extensions; reason should have min length (10 chars).