A boilerplate MCP server with Stripe integration for paid tools and Google/Github authentication, deployed on Cloudflare Workers
Six tools with basic schemas and descriptions, but significant gaps in parameter documentation and error handling. All tools have descriptions (10-60 chars, below the 194-char baseline), and input schemas are present but minimal. No tool annotations (readOnlyHint, destructiveHint, idempotentHint). Payment-related tools (onetime_add, subscription_add, metered_add) lack documentation of payment prerequisites, error states, and recovery paths. Naming is verb-noun compliant but generic (add, calculate). No output schemas documented. Error handling absent from visible code.
This tool adds two numbers together.
This tool performs a calculation on two numbers.
This tool checks for active subscriptions and one-time purchases for the logged in user's Stripe customer ID.
Adds two numbers together for metered usage.
Adds two numbers together for one-time payment.
Adds two numbers together for paid subscribers.
Payment tool descriptions lack prerequisites and error guidance. 'Adds two numbers together for paid subscribers' does not explain what happens if user lacks subscription, how to recover, or what payment states are possible.
No output schemas documented. LLMs cannot infer what fields are returned (e.g., does add return {result: number} or {sum: number, timestamp: string}?). Breaks downstream tool chaining.
check_payment_history has empty input schema ({}). No parameters documented. Unclear if it requires user context, Stripe customer ID, or filters. LLM cannot determine when to call it.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | D | 59 | 2026-07-28+ | v2 |
No tool annotations (readOnlyHint, destructiveHint, idempotentHint). All tools marked READ_ONLY in metadata, but this is not reflected in tool definitions. LLMs cannot determine retry safety or side effects.
Descriptions are 35 - 60 characters, well below the 194-char baseline. Too brief to guide LLM selection. E.g., 'Adds two numbers together for metered usage' does not explain metered billing model, rate limits, or when to use vs subscription_add.