Tool Function Schema Designer

What it does

This skill automatically generates structured schemas and function definitions for databases, designed to be reliably invoked by Large Language Models (LLMs). It produces code in a format compatible with OpenAI's function calling schema, including TypeScript interfaces and Zod validation schemas. The goal is to simplify the process of creating robust tool integrations for AI agents that interact with data sources.

When to use it

• You need to define structured inputs and outputs for an LLM-powered agent interacting with a database.
• You want to ensure consistent data types and formats when passing information between your agent and backend functions.
• You are building custom tools that require precise parameter validation before execution.
• You're working on integrating multiple tools into a single AI agent workflow.

Key capabilities

• Generates OpenAI function calling schemas (including name, description, parameters).
• Creates corresponding TypeScript interfaces for type safety.
• Provides Zod schemas for runtime validation of tool parameters.
• Includes example implementation functions and a tool registry structure.

Example prompts

• "Create a schema for a function that searches product documentation, allowing filtering by category and version."
• "Generate a tool definition to create support tickets with fields for subject, description, and priority."
• "Design a function schema for retrieving user profiles, including optional parameters for sorting and pagination."

Tips & gotchas

• The generated schemas are designed for OpenAI's function calling format. Ensure your LLM is configured to utilize this format.
• This skill focuses on generating the schema and validation logic; you will need to implement the actual backend functionality (e.g., the search_documentation function in the example).