Python package for integrating OpenAPI tools with AI providers (Anthropic, OpenAI, LangChain). Loads tools from local folders or API, converts them to provider-specific formats, and handles tool execution.
Single tool 'GenerateOtpTool' has critical definition gaps. Description is present (56 chars, within range) but schema is severely malformed: phoneNumber typed as 'integer' (should be string for phone numbers), openv parameter lacks type definition entirely, and no output schema is documented. Tool name lacks action verb prefix. No error handling guidance. Parameter descriptions are minimal. This server fails basic quality standards for production use.
Generates an OTP (One-Time Password) for a given phone number
Tool name 'GenerateOtpTool' does not start with action verb. Should be 'generate_otp' following verb_noun convention. LLMs rely on verb-prefixed names to infer intent.
Input schema critically malformed: phoneNumber typed as 'integer' but phone numbers are strings (can contain +, -, spaces, extensions). openv parameter has no type definition at all. Both violate JSON Schema requirements.
No output schema documented. LLMs cannot plan downstream operations or extract OTP value. Tool description does not specify return format (is it a string? object with success/otp fields?).
Parameter 'openv' (environment variables containing secret) exposed as tool input parameter. Secrets must never be passed as parameters, they appear in logs and traces. Use server-side secret injection instead.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | F | 28 | 2026-07-28+ | v2 |
No error handling guidance. Example shell script shows validation ('Phone number is required') but tool definition provides no recovery hints. LLMs cannot self-correct on invalid input.