54 tools with consistent naming (verb_noun pattern: ha_*, esphome_*, nodered_*, vomehome_*) but critical gaps in schema visibility and parameter documentation. Tool descriptions are present but generic (avg ~60 chars, well below 194-char baseline). No visible input schemas in provided source, cannot verify parameter types, constraints, or output structures. Risk categorization (READ_ONLY, WRITE, DESTRUCTIVE) is helpful but does not substitute for formal schema documentation. Composition is sound (single responsibility per tool), but LLM-facing quality is undermined by missing parameter descriptions and undocumented output formats.

• Add explicit JSON Schema definitions for all 54 tools. Include type, description, enum constraints, and min/max for every parameter. Example: { 'entity_id': { 'type': 'string', 'description': 'Home Assistant entity ID (e.g., light.living_room)', 'pattern': '^[a-z_]+\.[a-z_0-9]+$' } }
• Expand tool descriptions to 80-150 chars. Include: (1) what the tool does, (2) when to use it vs similar tools, (3) key prerequisites. Example: 'Call a Home Assistant service (e.g., turn on a light). Use this after ha_list_services to see available services. Requires service domain and entity_id.'
• Document output schemas for all tools. Specify returned fields, types, and pagination. Example: 'Returns array of {entity_id: string, state: string, last_changed: ISO8601}. Paginated: limit=50, next_cursor in response.'
• Add confirmation/dry-run pattern for destructive tools (delete_*, remove_*). Implement a 'confirm' parameter or separate 'preview_delete_user' tool. Return what will be deleted before executing.
• Add error recovery guidance to tool descriptions. Example for ha_call_service: 'If service not found, call ha_list_services first. If entity not found, call ha_list_entities. If permission denied, check Home Assistant user role.'

Inferred effective spec: <=2025-11-25.

• Relies on Logging (deprecated) - log to stderr or use OpenTelemetry