tanstack-ai

Integrate a provider-agnostic AI SDK for streaming, tool calling, and multimodal outputs into JavaScript or TypeScript projects.

Updated Apr 14, 2026
One-click install
npx skills add https://github.com/iEnergyy/opsflow --skill tanstack-ai-ienergyy
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: tanstack-ai
Source: https://github.com/iEnergyy/opsflow/tree/main/.agents/skills/tanstack-ai
Command: npx skills add https://github.com/iEnergyy/opsflow --skill tanstack-ai-ienergyy

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Provider-agnostic, type-safe AI SDK for streaming, tool calling, structured output, and multimodal content to simplify integrating AI into apps without vendor lock-in.

Core Features & Use Cases

  • Streaming-first generation with chunked outputs and optional thinking tokens.
  • Tool calling with approval workflows and runtime adapter switching.
  • Structured output via Zod schemas for strongly typed AI responses.
  • Multimodal content support including text and images across React, Solid, and vanilla JS UIs.
  • Framework adapters and UI hooks (React, Solid, Vanilla) to accelerate integration and reuse.

Quick Start

Install the TanStack AI packages and initialize a client with your chosen adapters to start streaming and tool calls.

Frequently Asked Questions about tanstack-ai

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I integrate AI streaming into a JavaScript project without vendor lock-in?

Use a provider-agnostic AI SDK to integrate streaming into JavaScript projects, enabling chunked outputs and runtime adapter switching to avoid vendor lock-in. It provides modular adapters to connect your chosen provider seamlessly.

Can I generate structured output from AI responses using Zod schemas?

Yes, you can generate structured output using Zod schemas for strongly typed AI responses. The SDK validates AI outputs against your schema definitions, ensuring type safety and predictable data structures across your application.

Does the AI SDK support tool calling with approval workflows?

The AI SDK supports tool calling with built-in approval workflows. You can execute tool calls dynamically while maintaining control over the process through runtime adapter switching and structured approval pipelines.

How do I build multimodal chat UIs across React and Solid?

Build multimodal chat UIs using framework adapters and UI hooks for React, Solid, and vanilla JS. The SDK supports multimodal content including text and images, accelerating integration and component reuse across frameworks.

What is the best way to add thinking tokens to streaming AI generation?

Streaming-first generation handles thinking tokens by processing chunked outputs during the stream. You can configure optional thinking tokens to capture intermediate AI reasoning directly within your streaming pipelines.

Does this AI SDK work with vanilla JavaScript or only TypeScript?

The SDK works with both JavaScript and TypeScript projects. It provides framework adapters and UI hooks for React, Solid, and vanilla JS UIs, ensuring type-safe integration while supporting vanilla JavaScript environments.