Four tools with explicit schemas and descriptions. Naming follows verb_noun pattern (list_, get_, create_). Descriptions are present but vary in quality (10-200 chars). Schemas are well-structured with enums and required fields. However, parameter descriptions are sparse or missing in several tools, and output schemas are not documented. Error handling is minimal. Token-budget optimization is evident but reduces discoverability.

• Document output schemas for create_poster and create_poster_variants. Specify: path (string), width (number), height (number), format (string), bytes (number), template (string), palette (object with bg/accent hex), htmlPath (string|null), preview (base64 string|null). This enables LLMs to chain results to downstream tools.
• Move field documentation into the create_poster schema itself. Instead of 'Fields: list_poster_templates', embed a nested object schema with per-template field definitions and required markers. Reduces discovery friction.
• Add format/pattern constraints to brandColor and palette parameters. E.g. brandColor: {type: 'string', pattern: '^#[0-9A-Fa-f]{6}$', description: 'Hex color code (e.g. #FF5733). A full palette is derived from it.'}
• Add error recovery guidance to tool descriptions. E.g. create_poster: 'If render fails, check that all required fields for the template are present. Call list_poster_templates to verify field names and requirements.'
• Specify numeric bounds for quality parameter: quality: {type: 'number', minimum: 1, maximum: 100, description: 'JPEG quality 1-100, default 92.'}. Prevents LLMs from passing invalid values like 150 or 0.