https://github.com/romanslack/simuverseframework_mk1
The framework powering the SimuVerse simulation.
https://github.com/romanslack/simuverseframework_mk1
Last synced: 9 months ago
JSON representation
The framework powering the SimuVerse simulation.
- Host: GitHub
- URL: https://github.com/romanslack/simuverseframework_mk1
- Owner: RomanSlack
- License: mit
- Created: 2025-03-16T21:25:51.000Z (over 1 year ago)
- Default Branch: main
- Last Pushed: 2025-04-11T20:57:40.000Z (about 1 year ago)
- Last Synced: 2025-04-11T21:29:53.158Z (about 1 year ago)
- Language: Python
- Size: 43 KB
- Stars: 0
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# SimuVerse Framework
The framework powering the SimuVerse simulation.
## Setup
1. Create a `.env` file with your OpenAI API key:
```
OPENAI_API_KEY=your_api_key_here
```
2. Install requirements:
```bash
pip install fastapi uvicorn openai python-dotenv
```
## Running the Framework
1. Start the main API server:
```bash
python main.py
```
This will start a FastAPI server on http://127.0.0.1:3000 that provides the agent interaction API.
2. Start the logs viewer interface (in a separate terminal):
```bash
python logs_viewer.py
```
This will start a separate web interface on http://127.0.0.1:3001 that shows logs for all agents in a nice, user-friendly interface.
## API Endpoints
### Main API (port 3000)
- `POST /generate` - Generate a response from an agent
- `GET /api/logs` - Get all logs in JSON format
- `GET /api/logs/{agent_id}` - Get logs for a specific agent in JSON format
- `POST /clear_logs` - Clear all logs
- `POST /reset` - Reset the entire system (clear sessions and logs)
### Logs Viewer (port 3001)
- `GET /` - Web interface for viewing logs
- `POST /clear_logs` - Clear all logs
## Logs Viewer Features
- View logs for all agents in a clean, user-friendly interface
- Switch between standard and chat views
- Auto-refresh functionality to see logs in real-time
- Clear logs with a single click
- Mobile-responsive design
## How Logs Work
1. The system logs all agent interactions to an `agent_logs.json` file
2. Logs are organized by agent ID
3. Each log entry includes:
- Timestamp
- Event type (user_input, generating_response, response, validation_failure)
- Detailed information specific to the event type
4. Logs persist between server restarts but can be cleared via the API or UI
## Example Agent Interaction
```python
import requests
# Generate a response from an agent
response = requests.post(
"http://127.0.0.1:3000/generate",
json={
"agent_id": "agent1",
"user_input": "Hello, how are you?",
"system_prompt": "You are a helpful assistant."
}
)
print(response.json())
```
## Model
Current model: gpt-4o-mini-2024-07-18