@sylphlab/tools-adaptor-vercel

Integrate SylphLab Tools with the Vercel AI SDK seamlessly!

This package provides adapter functions to convert tool definitions created using @sylphlab/tools-core into the format expected by the Vercel AI SDK (ai package), specifically for use with Large Language Models (LLMs) that support tool/function calling.

Purpose

The Vercel AI SDK offers a powerful way to integrate generative AI features into applications. When using LLMs that support tool calling (like OpenAI's GPT models), the SDK requires tool definitions in a specific format (based on JSON Schema).

If you define your tools using @sylphlab/tools-core, this adapter bridges the gap by converting those definitions into the structure the Vercel AI SDK needs to pass to the LLM and to handle tool execution responses.

Key Features

• Automatic Conversion: Transforms SylphTool definitions into the tools format compatible with the Vercel AI SDK's generateText or similar functions.
• Schema Translation: Leverages zod-to-json-schema to convert Zod schemas (used in @sylphlab/tools-core) into the JSON Schema format required by the Vercel AI SDK and underlying LLMs.
• Metadata Mapping: Ensures tool names, descriptions, and parameter details are correctly formatted.
• Simplifies AI Integration: Makes it easier to expose your existing SylphLab tools to LLMs via the Vercel AI SDK without manual definition rewriting.

Installation

This package is primarily intended for internal use within the mcp monorepo. If you are building an application using the Vercel AI SDK and want to leverage tools defined in this repository, add it as a dependency:

# From the root of the monorepo
pnpm add @sylphlab/tools-adaptor-vercel --filter <your-vercel-ai-app-package>

Usage (Conceptual)

You would typically use this adapter when preparing the tools option for a Vercel AI SDK function call:

import { generateText } from 'ai';
import { openai } from '@ai-sdk/openai'; // Or your preferred provider
import { adaptToolToVercel } from '@sylphlab/tools-adaptor-vercel';
import { filesystemTools } from '@sylphlab/tools-filesystem'; // Assuming this exports an array of SylphTool definitions

// Adapt SylphTool definitions for Vercel AI SDK
const vercelAiTools = {};
filesystemTools.forEach(tool => {
  vercelAiTools[tool.name] = adaptToolToVercel(tool);
});

async function main() {
  const { toolResults, text } = await generateText({
    model: openai('gpt-4-turbo'),
    prompt: 'Create a directory named "test-dir" and then list the contents of the current directory.',
    tools: vercelAiTools, // Pass the adapted tools here
  });

  // Handle toolResults if any tools were called...
  // Process the generated text...
  console.log(text);
}

main();

(Note: The actual implementation depends on your specific Vercel AI SDK setup and how you handle tool execution.)

Dependencies

• ai: The Vercel AI SDK.
• @sylphlab/tools-core: Provides the base SylphTool definition format.
• zod: Used for schema definition.
• zod-to-json-schema: Used for converting Zod schemas to JSON Schema.

Developed by Sylph Lab.