Loading ToolBench
Product Hunt MCP — wraps the Product Hunt GraphQL API v2 (api.producthunt.com). Surfaces daily product / startup launches: what launched today or recently, trending products, and full details for any individual launch.
Product Hunt MCP has clear, well-structured tool definitions with proper schemas and descriptions. Both tools follow verb_noun naming (top_launches, get_post) and include detailed parameter documentation. However, output schemas are not formally documented, error handling lacks recovery guidance, and parameter descriptions contain example values (anti-pattern). The dual API key model is well-explained but introduces complexity. Tool descriptions are solid (150-200 chars, within baseline), and all parameters have types and descriptions. No critical gaps, but missing output schema documentation and actionable error messages prevent a higher score.
Get full details for a single Product Hunt launch by its slug — description, makers, topics, and vote/comment counts. Use after top_launches to dig into a specific new product / startup launch. Example: get_post({ slug: "notion" })
Get today's and recent top Product Hunt launches — new product and startup launches. Use this to see what launched today/recently or which products are trending. Returns name, tagline, vote/comment counts, URLs, and topics. Example: top_launches({ first: 10, order: "RANKING" })
Output schemas not documented. LLMs cannot plan downstream calls or extract fields without knowing the response structure. Both tools return structured objects (launches array, post object) but this is inferred from code, not declared in tool definitions.
Error responses lack recovery guidance. phQuery returns {error, message} but does not tell the LLM what to do next (e.g., 'Try with your own API key' or 'Check the slug format'). Bare error codes force the agent to guess.
Parameter descriptions contain example values ('notion', 'RANKING'). LLMs tend to reuse examples literally rather than adapting to context. Use enums and format constraints instead.
No pagination support. top_launches caps at 20 results, but no offset/cursor or total_count returned. If Product Hunt has >20 launches, agents cannot discover older ones.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | C | 67 | 2026-07-28+ | v2 |
API key handling via _apiKey parameter is a security anti-pattern. Credentials in tool parameters leak into logs and traces. Use server-side secret injection (environment variables or vault) instead.