MCP server: generate and verify TOTP and HOTP one-time passwords.
Three well-defined tools with complete input schemas and clear descriptions. All tools have proper JSON Schema with type constraints and enums. Descriptions are concise (50 - 100 chars) and explain the RFC standard and use case. However, parameter descriptions are sparse, most lack guidance on format, constraints, or when to use them. Output schemas are not documented, forcing LLMs to infer structure. No error recovery guidance. Tool names are action-oriented (totp, hotp, uri) but lack verb prefixes, which is acceptable for domain-specific tools. No tool annotations (readOnlyHint, etc.) despite all being read-only operations.
Generate an HOTP code (RFC 4226). Caller supplies the counter.
Generate a TOTP code (RFC 6238). Secret is base32. Default period=30, digits=6, algo=SHA1.
Build an otpauth:// URI suitable for QR-coding into Google Authenticator etc.
Output schemas not documented. LLMs cannot infer what fields totp returns (code, counter, period) or hotp/uri structure without reading source code.
Parameter descriptions missing or minimal. 'secret' has no description explaining base32 encoding requirement. 'at' lacks guidance on millisecond vs second precision. 'period', 'digits', 'algorithm' lack context on valid ranges and defaults.
No tool annotations despite all tools being read-only. Adding readOnlyHint=true would signal to agents these are safe to call without side effects.
Error handling returns generic 'otp failed: <message>' without recovery guidance. If base32 decoding fails, LLM has no hint to validate secret format or try search_secrets().
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | D | 58 | 2026-07-28+ | v2 |
Tool names lack action verbs. 'totp', 'hotp', 'uri' are nouns/acronyms. Renaming to 'generate_totp', 'generate_hotp', 'build_uri' would improve LLM intent inference.