How ari-111/honeypot works
This is a FastAPI scam-detection honeypot: an incoming message is validated and rate-limited, analyzed for scam intent, answered in a selected persona when suspicious, and mined for intelligence; qualifying sessions are reported to GUVI. The graph follows relationships visible in excerpts, the SOURCE INDEX, or README. Engagement strategy and several documented API capabilities are included, but unsampled implementation details and their connections are not inferred.
Main entry points: routes.py Honeypot endpoints · main.py FastAPI app · agent_manager.py Agent manager · callback_handler.py GUVI callback handler
Components 17
API and sessions
- FastAPI app
main.py - Honeypot endpoints
api/routes.py - Rate limiter
api/middleware.py - Request and response models
models/session.py - Session state
api/routes.py
Detection and intelligence
- Scam detector
core/scam_detector.py - Intelligence extractor
core/intelligence_extractor.py - Data validators
utils/validators.py
Agent conversation
- Agent manager
core/agent_manager.py - Persona selection
core/persona_engine.py - Persona profiles
personas - Engagement strategy
core/engagement_strategy.py
Reporting and configuration
- OpenAI service
- GUVI callback handler
utils/callback_handler.py - Runtime configuration
config.py
Other components
- API caller
- GUVI evaluation endpoint
Connections 19
- API caller → FastAPI app: sends requests
README.md - FastAPI app → Honeypot endpoints: registers routes
main.py - FastAPI app → Rate limiter: installs limiter
main.py - FastAPI app → Request and response models: validates fallback requests
main.py - Honeypot endpoints → Request and response models: uses API models
api/routes.py - Honeypot endpoints → Session state: manages sessions
api/routes.py - Honeypot endpoints → Scam detector: detects scam intent
api/routes.py - Honeypot endpoints → Agent manager: generates replies
api/routes.py - Honeypot endpoints → Intelligence extractor: extracts intelligence
api/routes.py - Honeypot endpoints → Persona selection: selects persona
api/routes.py - Persona selection → Persona profiles: loads profiles
core/persona_engine.py - Agent manager → OpenAI service: requests completions
core/agent_manager.py - Agent manager → Runtime configuration: reads model settings
core/agent_manager.py - Rate limiter → Runtime configuration: reads rate limits
api/middleware.py - Honeypot endpoints → GUVI callback handler: triggers reports
api/routes.py - GUVI callback handler → GUVI evaluation endpoint: posts findings
utils/callback_handler.py - GUVI callback handler → Runtime configuration: reads callback settings
utils/callback_handler.py - Engagement strategy → Agent manager: guides conversation
- Intelligence extractor → Data validators: validates extracted data