MCP server for Renpho smart scale body composition data
Server has 9 tools with basic descriptions and some parameter schemas. However, critical gaps undermine quality: 4 tools (get_current_user, get_scale_users, get_latest_measurement, get_body_composition, refresh_data, health_check) have ZERO input parameters documented, no schema visible. Descriptions are present but generic (avg ~80 chars, below 194-char baseline). No output schemas documented. Error handling logs errors but provides no recovery guidance to LLMs. All tools are READ_ONLY, reducing risk, but composition and chaining support are weak. STDIO transport caps score at 50 maximum; only logging feature present.
Get detailed body composition analysis with health classifications
Get current Renpho user information
Get the most recent body composition measurement from Renpho scale
Get historical body composition measurements
Get linked Renpho scale-user IDs and table mappings discovered from the account
How many recent days to inspect for hidden or delayed measurements (default: 7). Also reports every raw device/count category on the account (including ones this server does not read yet) and the device model behind each recent measurement.
6 of 9 tools have NO input schema (get_current_user, get_scale_users, get_latest_measurement, get_body_composition, refresh_data, health_check).
No output schemas documented for any tool. LLMs cannot plan downstream calls or extract required fields (e.g., user_id, measurement_id) without knowing response structure.
Error handling logs errors but provides no recovery guidance. Catch blocks log to file but throw raw errors without actionable messages for LLMs (e.g., 'User not found. Try search_users()' or 'API rate limited. Retry in 60s').
Inferred effective spec: <=2025-11-25.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | D | 51 | <=2025-11-25 | v2 |
Get weight trend analysis over a specified period
Check the health status of the Renpho MCP server and API connection
Clear Renpho auth/measurement caches and re-fetch account state. Useful after taking a new measurement or after the mobile app syncs.
Tool descriptions are generic and under 100 chars (avg ~70). Baseline is 194 chars. Descriptions lack WHEN to use, prerequisites, or dependencies. E.g., 'Get current Renpho user information' does not explain when to call it vs get_scale_users.
No tool composition or chaining support documented. Response fields (e.g., user_id, measurement_id) are not explicitly listed, so LLMs cannot know what IDs to pass to downstream tools.