Loading ToolBench
A Model Context Protocol (MCP) server that provides search and crawl functionality using Search1API
Five tools with complete input schemas and descriptions. All tools are read-only web operations with proper annotations. Descriptions are clear and action-oriented (avg ~120 chars), exceeding the 10-char minimum. Parameters are well-typed with enums for constrained inputs (search_service, time_range). Output schemas are not explicitly documented in the source, which is a notable gap. Error handling is present but generic, no recovery guidance or actionable error messages visible in tool definitions. Tool composition is sound: search/news/crawl/sitemap/trending each have a single responsibility. Parameter naming is consistent (query, max_results, url). No secrets exposed as parameters.
Extract the readable title and full text from a public URL supplied by the user or returned by search.
Search current news when the user asks about recent events, announcements, or coverage. Returns citable article results with title, URL, snippet, and optional extracted text.
Search the live public web when the user needs current information, sources, or research. Returns citable results with id, title, URL, and text. Pass a result URL to crawl to retrieve the full page.
Discover related public links from a page or domain when the user wants to explore a site's structure or available pages.
List currently trending repositories or stories from GitHub or Hacker News when the user asks what is popular now.
Output schemas not documented. Tools return results but LLMs cannot see expected response structure (fields, types, pagination). This forces LLMs to guess at response shape and risks parsing errors.
No error recovery guidance in tool descriptions. Tools may fail (invalid URL, rate limit, network error) but descriptions do not explain what to do next or how to self-correct.
Parameter 'crawl_results' and 'max_results' lack explicit guidance on cost/credit implications. Description mentions 'adds 1 credit' but does not explain the billing model or when to use crawl vs search alone.
Sitemap and crawl tools accept arbitrary URLs without documented validation or safety constraints. No mention of rate limits, timeout behavior, or handling of large/malicious pages.
Inferred effective spec: 2026-07-28+.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | B | 71 | 2026-07-28+ | v2 |
No pagination or result-limiting guidance for search/news tools. While max_results is capped at 50, no mention of whether results are ranked by relevance or how to iterate through large result sets.