How q-ai-adhd/q-ai_ai works
This is a small FastAPI application for Korean technical-interview question generation and answer feedback, using prompt templates and a locally loaded llama.cpp model. The principal workflows are POST question request → question prompt → model generation → question response, and POST feedback request → feedback prompt → model generation → feedback response. The README only documents llama-cpp-python installation; workflow links are evidenced by main.py. The feedback schema excerpt does not define the feedback classes imported there, so its precise request/response fields are uncertain.
Main entry points: main.py FastAPI Handlers · question_prompt.py Question Prompt · feedback_prompt.py Feedback Prompt · questSchema.py Question Schema
Components 7
API
- FastAPI Handlers
main.py
Prompt Templates
- Question Prompt
prompt/question_prompt.py - Feedback Prompt
prompt/feedback_prompt.py
Request Contracts
- Question Schema
schemas/questSchema.py - Feedback Schema
schemas/feedbackSchema.py
Other components
- API Client
- Local llama.cpp
Connections 7
- API Client → FastAPI Handlers: posts requests
main.py - FastAPI Handlers → Question Schema: validates and shapes
main.py - FastAPI Handlers → Feedback Schema: uses feedback types
main.py - FastAPI Handlers → Question Prompt: builds question prompt
main.py - FastAPI Handlers → Feedback Prompt: builds feedback prompt
main.py - FastAPI Handlers → Local llama.cpp: generates output
main.py - FastAPI Handlers → API Client: returns responses
main.py