AI Models and Prompt Manager

About - Help

This web application serves as a unified hub for managing AI tools and local language models. Users can add and organize web-based AI applications by URL (note: applications must support CORS for proper iframe embedding). The app seamlessly connects to local LLM instances running on LM Studio, allowing you to chat with your own models nd open code blocks in Livecodes.

Features include a built-in prompt library for saving and reusing frequently used prompts, and comprehensive settings management for API endpoints and model configurations. All settings and data are stored locally in your browser's localStorage, ensuring privacy and persistence across sessions. You can export and import your settings and prompts.

Key Features

Web AI Applications: Add and manage web-based AI application URLs with iframe integration
Local LLM Support: Connect to local LM Studio instances for private AI conversations
Prompt Management: Built-in library for saving and organizing frequently used prompts
Settings Persistence: All configurations saved locally in browser localStorage
Responsive Design: Mobile-friendly interface with collapsible sidebar navigation
Multiple Endpoints: Support for various AI model APIs and services

Important Notes

CORS Requirement: Web applications must support Cross-Origin Resource Sharing (CORS) to load properly in iframes
Local Storage: All data is stored locally - clearing browser data will remove your settings and prompts
LM Studio Connection: Ensure LM Studio is running and accessible at the configured URL for local AI features
Privacy: No data is sent to external servers - everything runs locally in your browser

Getting Started

Configure your LM Studio URL in Settings if using local AI models
Add web AI applications through the Settings panel
Create and organize prompts in the Prompts library
Switch between different AI tools using the dropdown in the header
Backup your settings and prompts (export and import json)

Settings - Manage APIs

LM Studio Server

The local server is reachable at this address

LM Studio Models

Switching models may load a new model, increasing memory usage and causing a delay. Models unload after 60 minutes of inactivity (Idle TTL) unless Auto-Evict unloads prior models first.

Supported API Endpoints

GET /api/v0/models - List available models
GET /api/v0/models/{model} - Get info about a specific model
POST /api/v0/chat/completions - Chat Completions (messages → assistant response)
POST /api/v0/completions - Text Completions (prompt → completion)
POST /api/v0/embeddings - Text Embeddings (text → embedding)

New Application

Name	URL	Category	Actions