How akhilchiluvari/project-jarvis works
JARVIS OS is represented here as a FastAPI chat-streaming backend. The visible workflow accepts a chat request, passes its message and session ID to the agent, and emits agent tokens and tool-status events as server-sent events. The backend also checks Redis at startup, but the excerpt does not show chat state being stored there. The frontend and agent implementation are not excerpted, so their internal behavior and wiring to external model services cannot be confirmed.
Main entry points: main.py Chat stream API · frontend Frontend app · agent.py Agent
Components 5
Chat runtime
- Frontend app
apps/frontend - Chat stream API
services/core-backend/main.py - Agent
services/core-backend/agent.py
State infrastructure
- Redis
Other components
- Chat client
Connections 4
- Chat client → Frontend app: uses
- Frontend app → Chat stream API: submits chat
- Chat stream API → Agent: runs agent
services/core-backend/main.py - Chat stream API → Redis: checks connection
services/core-backend/main.py