Three tools with basic schemas and descriptions, but significant gaps in parameter documentation, output schema clarity, and error handling. Tool names are action-oriented (analyze_, export_, capture_) but descriptions lack depth on when/why to use each tool. Parameters present but many lack detailed constraints or format guidance. No documented output schemas. Error handling is minimal, Python subprocess errors are caught but not actionable for LLMs. Missing idempotency guidance and confirmation patterns for destructive operations (export_components writes files).

• Document output schema for analyze_image: return object with 'components' array, each containing {id, name, category, description, reusable, html}. Include example structure.
• Add format constraints to parameters: image_path max 500 chars, output_dir must be writable directory, selected_ids array 1-100 items. State these in descriptions.
• Make export_components require 'image_path' explicitly and document the dependency: 'image_path must match the screenshot analyzed by analyze_image to locate the components.json file.'
• Implement error categorization: return {error_type: 'file_not_found'|'invalid_format'|'api_error'|'timeout', message: '...', recovery: 'Try...'}. Guide LLM on retry vs user action.
• Add dry_run parameter to export_components (default false) so agents can preview exports before writing. Return list of files that would be created without side effects.