Loading ToolBench
MCP server providing access to Product Hunt data including product listings, reports, GitHub Trending reports, and stock news. Operates in HTTP JSON-RPC mode on port 8080.
The server defines 9 tools with explicit schemas and descriptions in Chinese. All tools have inputSchema objects with proper structure. Naming follows verb_noun conventions (get_*, search_*). However, descriptions lack context about dependencies, error conditions, and when to use each tool vs. similar ones. Parameter descriptions are present but often minimal (e.g., 'The product quantity limit' without mention of actual constraints or why limits exist). Output schemas are not documented, the code references response filtering (filter_product_fields) but doesn't expose a schema to the LLM. Error handling is absent from tool specifications. Parameter types are complete but some descriptions could be more actionable (e.g., 'days_ago' should explain what happens on invalid dates). The tool set shows reasonable cohesion around Product Hunt data retrieval but lacks the sophistication of production-grade agent tools in error recovery and chaining.
获取 GitHub Trending 日报。可以指定日期获取特定日期的日报，不指定则返回最新日报。
获取最新的 Product Hunt 产品列表。返回数据只包含中文内容（tagline_cn, description_cn）。默认获取今天的数据，可以通过 days_ago 参数指定获取几天前的数据。
获取最新的 Product Hunt 每日报告。报告包含产品分析和趋势总结。
获取最新的美股科技股票资讯。自动处理周末不开盘的情况，返回最近交易日的股票新闻。
根据指定日期获取 Product Hunt 产品列表。返回数据只包含中文内容（tagline_cn, description_cn）。日期格式为 YYYY-MM-DD，例如：2024-03-15
根据指定日期获取 Product Hunt 每日报告。日期格式为 YYYY-MM-DD
Output schemas not documented. LLM cannot infer which fields are returned, breaking tool chaining and context planning. Code filters responses (filter_product_fields removes 'tagline', 'description') but LLM has no specification of actual output structure.
No error recovery guidance in tool descriptions. When a date is invalid or products are not found, no description tells the LLM what to do next (retry with a different date, try search instead, etc.).
Inferred effective spec: <=2025-11-25.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-22 | D | 53 | <=2025-11-25 | v2 |
| 2026-03-09 | F | 0 | - | v1 |
获取指定日期范围内的所有报告
获取指定日期投票数最多的热门产品。返回数据只包含中文内容（tagline_cn, description_cn）。
搜索 Product Hunt 产品。支持中英文关键词搜索，返回数据只包含中文内容（tagline_cn, description_cn）。
Parameter descriptions are minimal and lack actionable constraints. E.g., 'days_ago' has no explanation of edge cases (what if days_ago exceeds available data history?). 'limit' parameters describe the constraint but not why it exists or what happens if omitted.
Tool set includes overlapping responsibilities without clear disambiguation. 'get_top_products' could be confused with 'get_latest_products' filtered by votes. Descriptions should explain when to use each (e.g., 'Call get_top_products when user asks for trending/popular items; call get_latest_products for newest launches regardless of popularity').
Date format validation via regex pattern ^\d{4}-\d{2}-\d{2}$ is present in schema, but descriptions do not clarify timezone handling, what 'today' means (UTC? local?), or whether historical dates beyond a certain window are available.
No pagination or result limiting guidance. While 'limit' parameters exist, descriptions do not explain what happens if results exceed limit, whether results are sorted, or how to retrieve next page. Without pagination docs, LLM cannot efficiently iterate through large result sets.