MCP server for OpenWebTrack analytics platform, providing tools to query website analytics, breakdowns, records, and revenue data via authenticated API keys
Server exposes 4 analytics tools with complete input schemas and descriptions. All tools are READ_ONLY and properly scoped. However, descriptions lack depth and guidance for LLM selection. Parameter descriptions are minimal (e.g., 'Optional start date' without format hints). Output schemas are not documented, LLMs cannot predict response structure for downstream chaining. No error handling guidance or recovery paths documented. Tool names follow verb_noun convention but lack specificity (analytics_breakdown is vague about what 'breakdown' returns).
Break down one accessible website by pages, referrers, countries, browsers, devices, event names, or campaigns.
Get headline traffic, engagement, revenue, and top-page analytics for one accessible website.
Retrieve recent records from one accessible website. Supports pageviews, events, sessions, payments, and visitors; results are capped at 100.
List every website available to this account, including sites it owns and sites shared with it.
Output schemas not documented. LLMs cannot predict response structure (field names, types, nesting) for downstream tool chaining or data extraction.
Parameter descriptions lack format hints and constraints. 'Optional start date' does not specify ISO 8601, epoch, or natural language format. LLMs will guess and pass invalid formats.
No error handling guidance. If a websiteId is invalid or user lacks access, no recovery path is documented. LLMs cannot self-correct.
Tool descriptions lack context for LLM selection. 'Get headline traffic...' does not explain when to use analytics_overview vs analytics_breakdown vs analytics_records. LLMs will guess.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | C | 65 | 2026-07-28+ | v2 |
No pagination or result limit guidance in descriptions. analytics_records caps at 100 but does not explain how to fetch older records or handle large datasets.