Server provides 11 tools with complete input schemas and tool annotations (readOnlyHint, openWorldHint). All tools have descriptions (avg 85 chars, within baseline 34-392 range). However, descriptions are minimal and lack actionable context for LLM selection. Parameter descriptions are present but terse. No output schemas documented. Error handling not visible in source. Tool names follow verb_noun pattern but lack specificity, 'youtube-to-markdown' vs 'convert-youtube-video-to-markdown' would be clearer. No pagination, batching, or composition guidance evident.

• Document output schemas for all tools. Specify: does youtube-to-markdown return {markdown: string, transcript?: string, metadata?: {...}}? Does pdf-to-markdown return {markdown: string, page_count: number, has_images: boolean}? LLMs need this to plan downstream calls.
• Expand tool descriptions to 100-200 chars with context. E.g., 'Convert a PDF file to markdown. Use this for documents, reports, and forms. Returns structured markdown with headings, tables, and lists. Fails on scanned PDFs without OCR, use image-to-markdown for those.'
• Add parameter constraints and examples to descriptions. E.g., 'filepath: Absolute path to local file (e.g., /home/user/report.pdf). Supports PDF, DOCX, XLSX, PPTX. Max 100MB. Returns error if file not found or unsupported format.'
• Add error recovery guidance. E.g., 'If conversion fails due to unsupported format, try get-markdown-file to read a pre-converted version, or use webpage-to-markdown if the file is hosted online.'
• Consider renaming tools for clarity: 'extract-youtube-transcript', 'convert-pdf-to-markdown', 'scrape-webpage-to-markdown' to help LLMs distinguish intent and avoid conflation with similar tools.