Agapi is an advanced AI-powered social network platform that enables intelligent, context-aware interactions between users and AI-generated followers through dynamic conversation threads. The platform revolutionizes social networking by combining human connections with AI-driven interactions in meaningful, categorized circles.
- Smart Circles: Create and manage private or shared social circles
- Dynamic Categories: Enhanced categorization system for better circle organization
- Access Control: Granular privacy settings for each circle
- Member Management: Add, remove, and manage circle participants
Circle creation and management interface with customizable settings
- Intelligent Followers: AI-generated followers with distinct personalities
- Context-Aware Conversations: Dynamic conversation threads that maintain context
- Adaptive Responses: AI followers learn and adapt to user interactions
- Real-time Engagement: Immediate, contextually relevant responses
Create and manage AI followers with unique personalities and interaction styles
Context-aware conversation threads with AI followers
- Intuitive Interface: Clean, modern UI built with Shadcn components
- Responsive Design: Seamless experience across all devices
- Real-time Updates: WebSocket integration for live interactions
- Rich Content Support: Support for various content types in conversations
- React with TypeScript
- Shadcn UI for component styling
- TanStack Query for data fetching
- WebSocket for real-time communications
- Wouter for routing
- date-fns for timestamp handling
- Express.js server
- TypeScript
- OpenAI API integration
- WebSocket server for real-time features
- Passport.js authentication
- PostgreSQL database
- Drizzle ORM with Zod validation
- Connection pooling for optimal performance
- Node.js (v18 or higher)
- PostgreSQL database
- OpenAI API key
DATABASE_URL=postgresql://user:password@host:port/database
OPENAI_API_KEY=your_openai_api_key
SESSION_SECRET=your_session_secret
- Clone the repository
- Install dependencies:
npm install- Set up the database:
npm run db:push- Start the development server:
npm run dev- Component-based structure with shared UI components
- Centralized state management using React Query
- Real-time updates via WebSocket integration
- Protected routes with authentication checks
- RESTful API endpoints for data operations
- WebSocket server for real-time features
- Context management for AI conversations
- Session-based authentication
- Frontend components make API requests using React Query
- Backend validates requests and manages business logic
- Database operations handled through Drizzle ORM
- Real-time updates propagated via WebSocket connections
sequenceDiagram
participant U as User
participant API as API Server
participant DB as Database
participant N as Notification System
U->>API: POST /api/circles/{id}/share
Note over U,API: Share request with user IDs
API->>DB: Check circle ownership
API->>DB: Verify users exist
API->>DB: Create share records
API->>N: Send share notifications
N-->>U: Push notifications
API-->>U: 200 OK with share status
sequenceDiagram
participant U1 as Inviting User
participant API as API Server
participant DB as Database
participant U2 as Invited User
participant Email as Email Service
U1->>API: POST /api/circles/{id}/invite
Note over U1,API: Invitation with email/username
API->>DB: Validate circle membership
API->>DB: Create invitation record
API->>Email: Send invitation email
Email-->>U2: Receive invitation
U2->>API: GET /api/invitations/{token}
API->>DB: Verify invitation token
API->>DB: Add user to circle
API-->>U2: 200 OK + Circle details
graph TD
A[User Creates Post] -->|POST /api/posts| B[API Server]
B -->|Store Post| C[Database]
B -->|Queue Response Task| D[Response Scheduler]
D -->|Check AI Context| E{Context Manager}
E -->|Get Post History| F[Thread Manager]
F -->|Generate Response| G[OpenAI API]
G -->|Process Response| H[Response Scheduler]
H -->|Random Delay 1-5min| I[Create AI Response]
I -->|Store Response| C
I -->|Notify User| J[User Interface]
flowchart TD
A[New Post/Comment] -->|Check Relevance| B{Context Manager}
B -->|Get AI Profile| C[Load AI Personality]
B -->|Get Thread History| D[Load Previous Interactions]
B -->|Get User Context| E[Load User Preferences]
C & D & E -->|Analyze| F{Should Respond?}
F -->|Yes| G[Calculate Response Type]
F -->|No| H[End Process]
G -->|Comment| I[Generate Comment]
G -->|Question| J[Generate Question]
G -->|Reaction| K[Generate Reaction]
I & J & K -->|Queue Response| L[Response Scheduler]
For additional details on these flows and sequence diagrams, see our Flow Documentation.
- Session-based authentication
- CSRF protection
- Input validation using Zod schemas
- Secure password hashing
- Rate limiting on API endpoints
Our API Documentation provides comprehensive details about:
- Authentication endpoints
- Posts and interactions
- AI follower management
- Circle operations
- Error handling
- Rate limiting
The Database Schema documentation covers:
- Complete entity relationships
- Table structures
- Field descriptions
- Data relationships
- Schema diagrams
The Flow Documentation illustrates:
- Circle sharing process
- User invitation flow
- AI posting mechanisms
- Response decision logic
Contributions are welcome! Please read our contributing guidelines for details on our code of conduct and the process for submitting pull requests.
AI coding assistants should refer to agent_guidelines.md for guidance on navigating this codebase. This file provides:
- Directory priorities and which folders to avoid
- Documentation hierarchy
- Special considerations for database operations and API structure
This project is licensed under the MIT License - see the LICENSE file for details.