How asadmulani5/sentinelqa works
SentinelQA is a web-site QA scanning app: a user submits a URL, the backend fetches and extracts site metadata, an LLM generates an assessment, and the frontend presents the score and findings. The README describes unrelated portfolio projects, so this map follows the sampled SentinelQA source. The crawler is an HTTP/HTML metadata extractor; no browser automation or persistent store is shown.
Main entry points: agent.py Agent orchestration · App.jsx Scan and results UI · main.py Scan API
Components 8
Frontend
- Scan and results UI
frontend/src/App.jsx
Backend API
- Scan API
backend/main.py
Scan Agent
- Agent orchestration
backend/agent.py - Site crawler
backend/agent.py - AI analysis
backend/agent.py
Other components
- Site owner
- Target website
- Groq LLM API
Connections 13
- Site owner → Scan and results UI: enters URL
- Scan and results UI → Scan API: POST /scan
- Scan API → Agent orchestration: runs scan
- Agent orchestration → Site crawler: crawls URL
- Site crawler → Target website: fetches HTML
- Site crawler → Agent orchestration: returns metadata
- Agent orchestration → AI analysis: requests analysis
- AI analysis → Groq LLM API: sends prompt
- Groq LLM API → AI analysis: returns completion
- AI analysis → Agent orchestration: returns assessment
- Agent orchestration → Scan API: returns report
- Scan API → Scan and results UI: returns JSON
- Scan and results UI → Site owner: shows findings