Element Node MCP exposes 12 CMS tools with Italian descriptions and reasonable naming (verb_noun pattern). However, critical gaps undermine production readiness: (1) Input schemas are present but minimal, most tools accept empty objects or single untyped parameters, lacking type constraints and validation guidance. (2) Descriptions are domain-specific but often assume CMS expertise; parameter descriptions are sparse or missing entirely. (3) No output schemas documented, LLMs cannot predict response structure for chaining. (4) Error handling is absent, no recovery guidance, no categorization of retryable vs fatal errors. (5) Destructive operations (site_import with strategy='replace', update_page, delete operations) lack confirmation/dry-run patterns. (6) No tool annotations (readOnlyHint, destructiveHint, idempotentHint) despite clear risk levels. Average tool score: 52/100.

• Add explicit output schemas to all tools. Document response fields, types, and pagination. E.g., list_pages should return {pages: [{id, title, slug, published}], total: number}.
• Enhance input schemas with type constraints: use enums for strategy (merge|replace), boolean for dryRun, URL format for upload_media_from_url.url, string patterns for page slugs.
• Add parameter descriptions for all inputs. E.g., 'id' → 'Page ID (from list_pages)', 'strategy' → 'merge: update only fields in blueprint; replace: delete all pages and recreate'.
• Implement dry-run support for site_import and update_page. Return validation errors without persisting changes. Document in descriptions: 'Pass dryRun=true to validate without writing.'
• Add tool annotations: mark site_export, list_pages, get_page, list_media, get_site_settings, widget_reference, widget_quirks with readOnlyHint=true; mark site_import, create_page, update_page, update_theme, upload_media_from_url with destructiveHint=true.