https://github.com/atomantic/portos-ai-toolkit
AI Provider system for PortOS apps
https://github.com/atomantic/portos-ai-toolkit
Last synced: 2 months ago
JSON representation
AI Provider system for PortOS apps
- Host: GitHub
- URL: https://github.com/atomantic/portos-ai-toolkit
- Owner: atomantic
- License: mit
- Created: 2026-01-20T23:49:20.000Z (5 months ago)
- Default Branch: main
- Last Pushed: 2026-03-31T05:01:44.000Z (3 months ago)
- Last Synced: 2026-03-31T07:39:20.268Z (3 months ago)
- Language: JavaScript
- Size: 617 KB
- Stars: 0
- Watchers: 0
- Forks: 0
- Open Issues: 1
-
Metadata Files:
- Readme: README.md
- Changelog: CHANGELOG.md
- License: LICENSE
Awesome Lists containing this project
README
# portos-ai-toolkit
Shared AI provider, model, and prompt template patterns for PortOS-style applications.
## Installation
```bash
npm install portos-ai-toolkit
```
## Features
- **Provider Management**: Support for CLI-based (Claude Code, Codex, etc.) and API-based (OpenAI-compatible) AI providers
- **Prompt Templates**: Reusable prompt template system with variable interpolation
- **Run History**: Track and manage AI run history with streaming support
- **React Components**: Ready-to-use React components and hooks for AI provider management
- **Express Routes**: Pre-built Express route handlers for provider, prompt, and run management
## Screenshot
## Usage
### Server-side
```javascript
import { createAIRoutes, ProvidersService, RunnerService } from 'portos-ai-toolkit/server';
import express from 'express';
const app = express();
// Initialize services
const providersService = new ProvidersService('./data');
const runnerService = new RunnerService(providersService);
// Mount AI routes
app.use('/api', createAIRoutes({ providersService, runnerService }));
```
### Client-side
```javascript
import { AIProviders, useProviders, createApiClient } from 'portos-ai-toolkit/client';
// Use the full-featured AIProviders page component
function ProvidersPage() {
return ;
}
// Or use hooks for custom implementations
function CustomComponent() {
const { providers, loading, refresh } = useProviders();
// ...
}
```
### Shared utilities
```javascript
import { PROVIDER_TYPES, DEFAULT_TIMEOUT } from 'portos-ai-toolkit/shared';
```
## Provider Types
### CLI Providers
Execute AI commands via CLI tools like Claude Code or Codex:
- Claude Code CLI (`claude`)
- Codex CLI (`codex`)
- Custom CLI tools
### API Providers
Connect to OpenAI-compatible APIs:
- LM Studio
- Ollama
- OpenAI
- Any OpenAI-compatible endpoint
## API Reference
### Server Exports (`portos-ai-toolkit/server`)
- `createAIRoutes(options)` - Create Express router with all AI routes
- `ProvidersService` - Manage AI provider configurations
- `RunnerService` - Execute prompts and manage runs
- `PromptsService` - Manage prompt templates
- Route handlers: `createProvidersRoutes`, `createRunsRoutes`, `createPromptsRoutes`
### Client Exports (`portos-ai-toolkit/client`)
- `AIProviders` - Full-featured provider management page component
- `ProviderDropdown` - Dropdown for selecting providers
- `useProviders()` - Hook for provider state management
- `useRuns()` - Hook for run history
- `createApiClient(baseUrl)` - Create API client instance
### Shared Exports (`portos-ai-toolkit/shared`)
- `PROVIDER_TYPES` - Provider type constants
- `DEFAULT_TIMEOUT` - Default timeout value
## License
MIT