cf-webmcp defines 5 tools with complete JSON schemas and reasonable descriptions. All tools are READ_ONLY and follow a consistent pattern (URL template + optional parameters). Naming is clear and verb-based (sitemap_filter, rss_feed, dom_extract, http_json, http_get). However, descriptions are brief (40-80 chars), lacking context on WHEN to use each tool or dependencies between them. Parameter descriptions exist but are minimal. No output schemas are documented. Error handling guidance is absent. Tool composition is weak, no clear guidance on which tool to chain after another.

• Expand tool descriptions to 100-200 chars. Include WHEN to use (e.g., 'Use sitemap_filter to discover all pages on a site before extracting content from each'), and WHAT to do next (e.g., 'Returns a list of URLs; pass each to dom_extract to extract content').
• Document output schemas for all tools. Specify field names, types, and whether results are paginated. E.g., sitemap_filter should return {urls: [string], total: number, has_more: boolean}.
• Add format/constraint guidance to parameter descriptions. E.g., 'url_template: URL with optional {{param}} placeholders (e.g., https://example.com/page/{{id}}). Placeholders are replaced with tool input parameters.' and 'max_results: 1-200 (default 20). Larger values may timeout.'
• Add error handling guidance to descriptions. E.g., 'May fail if the URL is unreachable, returns non-200 status, or content-type is not allowed. Retryable errors include timeout and 5xx; non-retryable include 404 and invalid content-type.'
• Clarify tool composition in descriptions. E.g., 'sitemap_filter discovers URLs; use dom_extract to extract content from each URL. http_get is a fallback for non-HTML content.'