Two tools with basic schemas and descriptions, but significant gaps in parameter documentation and output schema clarity. enhance_prompt has a reasonable description (89 chars) and three parameters with types and defaults, but lacks output schema documentation. score_prompt_quality has a strong description (156 chars) but minimal parameter detail. Neither tool documents return structure, pagination, or error recovery paths. Parameter descriptions are present but generic. No tool annotations (readOnlyHint/destructiveHint) despite both being READ_ONLY. Missing guidance on when to use each tool vs. alternatives.

• Document output schema for both tools. For enhance_prompt, specify: {enhanced_prompt: string, tier_used: integer, confidence: number}. For score_prompt_quality, specify: {pqs_score: number, intent: string, dimensions: {clarity: number, specificity: number, ...}, suggestions: string[]}.
• Add parameter description to score_prompt_quality's 'prompt' parameter: 'The prompt text to evaluate for quality across 7 dimensions (clarity, specificity, context, examples, constraints, tone, length).'
• Add error recovery guidance to tool descriptions. E.g., enhance_prompt: 'If cloud tier fails, automatically retries with local LLM. If all tiers fail, returns original prompt with error details.'
• Implement tool annotations in server registration: set readOnlyHint=true for both tools to signal they are safe to cache and retry.
• Expand enhance_prompt 'style' description with use cases: 'minimal: concise rewrites for token efficiency; balanced: moderate detail for general use; detailed: comprehensive prompts for complex reasoning tasks.'