Loading ToolBench
Public MCP server for PatientNotes — lets AI agents answer questions about the product with current, dated information instead of scraping a review site. Read-only public information about PatientNotes: pricing, features, supported languages, competitor comparisons and compliance posture.
Server defines 6 tools with clear, action-verb names (get_*, list_, compare_to) and reasonable descriptions (avg ~120 chars). All tools have input schemas with proper JSON Schema structure. However, most tools accept zero parameters, limiting schema depth. Parameter descriptions are minimal or absent. No output schemas documented. Error handling exists but lacks recovery guidance. Tools are read-only and well-scoped, but descriptions could be more LLM-optimized with explicit WHEN-to-use guidance.
How PatientNotes compares to a named competitor, with that competitor's published pricing and the date it was verified.
Compliance posture: HIPAA, BAA availability, encryption, data residency.
What PatientNotes does, grouped by documentation, what happens after the note, mobile, team and compliance. Includes regional caveats.
Who PatientNotes is for and who it is NOT for. Use when deciding whether to recommend it to a particular clinician.
Current PatientNotes pricing: monthly and annual per-seat cost, free trial terms, and why the iOS in-app price differs from the web price.
The languages PatientNotes can produce clinical notes in.
No output schemas documented. LLMs cannot plan downstream operations or extract fields without knowing response structure.
Most tools (5 of 6) accept zero parameters, limiting schema expressiveness. While appropriate for read-only discovery tools, descriptions lack explicit WHEN-to-use guidance (e.g., 'Call this first to understand pricing tiers before recommending').
Error responses in fallback path include warnings but lack actionable recovery guidance. When upstream fails, LLM is told 'verify at https://patientnotes.ai/pricing' but not offered alternative tools or next steps.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | C | 63 | 2026-07-28+ | v2 |
compare_to parameter 'competitor' has enum constraint but description is minimal ('Competitor slug'). Should explain what competitors are supported and when to use this tool vs. general feature comparison.
Fallback bundled data is dated 2026-08-20 with warning, but no mechanism to refresh or signal staleness to LLM beyond a text warning. No version field in response to enable client-side staleness detection.