LLM Proxy Server Integration

Stream real-time LLM responses in Android apps via Spring Boot proxy.

Updated May 10, 2026
One-click install
npx skills add https://github.com/nishchaymakkar/gemmabackned --skill llm-proxy-server-integration
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: LLM Proxy Server Integration
Source: https://github.com/nishchaymakkar/gemmabackned/tree/main
Command: npx skills add https://github.com/nishchaymakkar/gemmabackned --skill llm-proxy-server-integration

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This skill solves the challenge of connecting native Android applications to a Spring Boot LLM proxy backend to stream real-time AI responses, eliminating the need to build custom backend endpoints for chat completions, vision inputs, and web search functionality.

Core Features & Use Cases

  • Streaming Chat Completions: Send prompts to local or hosted LLMs via the proxy and receive real-time text chunks via Server-Sent Events for a smooth, ChatGPT-like typing effect in Android apps.
  • Vision Model Support: Send image inputs alongside text prompts to multimodal models like LLaVA or Moondream for image analysis use cases.
  • Optional Web Search Integration: Add a toggle to enable live web search for user queries, with the backend automatically injecting search context and source URLs into the model prompt.
  • Use Case: A developer building a native Android AI assistant can use this skill to connect to a local Gemma model, stream responses in real-time, add photo analysis features, and enable web search without writing any backend API code.

Quick Start

Use the LLM Proxy Server Integration skill to add real-time streaming AI chat to your native Android app by implementing the provided OkHttp SSE handler, ViewModel, and Jetpack Compose UI components.

Frequently Asked Questions about LLM Proxy Server Integration

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

FAQPage Schema
How do I stream LLM responses in an Android app using SSE?

You can stream LLM responses in an Android app by using an OkHttp SSE handler to connect to a Spring Boot proxy backend, receiving real-time text chunks via Server-Sent Events. This approach provides a smooth, ChatGPT-like typing effect in your native application UI.

How do I display streaming LLM tokens in Jetpack Compose?

Displaying streaming LLM tokens in Jetpack Compose involves managing the real-time text chunks via a ViewModel state. As the OkHttp SSE handler receives tokens from the Server-Sent Events stream, the ViewModel updates the state, allowing the Jetpack Compose UI to render the text incrementally.

Can I send image inputs to vision models through a Spring Boot proxy?

Yes, you can send image inputs alongside text prompts to multimodal vision models like LLaVA or Moondream through a Spring Boot proxy. This enables native Android applications to perform image analysis tasks by routing the data through the backend.

Does the LLM proxy backend support web search integration for chat completions?

Yes, the LLM proxy backend supports optional web search integration for chat completions. You can add a toggle to enable live web search, and the backend automatically injects search context and source URLs directly into the model prompt.

Do I need to write custom backend API code to connect my Android app to an LLM?

No, you do not need to write custom backend API code to connect an Android app to an LLM if you use a Spring Boot proxy backend. This setup eliminates the need to build custom endpoints for chat completions, vision inputs, and web search functionality.