Three tools with complete input schemas and descriptions, but significant gaps in parameter documentation and output schema clarity. Tool names follow verb_noun pattern (render_*, generate_*) which is good. However, descriptions lack LLM-optimized guidance on WHEN to use each tool vs alternatives, and many parameters lack descriptions or have generic ones. No output schemas documented. Error handling not visible in provided code. Baseline: avg tool description ~180 chars (acceptable range 10-1024), but parameter descriptions are sparse or missing entirely.

• Document output schema for each tool. Example: 'Returns {svg_path: string, jpeg_path: string, png_path?: string, ascii_output?: string, error?: string}' so LLM knows what to extract.
• Add parameter descriptions for all inputs. For 'source' in render_mermaid_text: 'Mermaid diagram syntax (flowchart, sequence, class, etc.). Max 100KB. Invalid syntax returns error with line number.'
• Clarify tool selection in descriptions. render_mermaid_text: 'Use when you have Mermaid syntax as a string.' render_mermaid_file: 'Use when diagram is in a .mmd/.mermaid file.' generate_source_diagrams: 'Use to auto-generate diagrams from source code analysis.'
• Replace 'string|array' type hints with explicit enum constraints. For 'langs': enum: ['typescript', 'python', 'java', 'go', ...] or accept comma-separated string with validation.
• Add error handling guidance. Example: 'If Mermaid syntax is invalid, returns {error: 'Syntax error at line X: ...'}. Suggest fixing the syntax and retrying.' Include timeout and file-not-found recovery hints.