Jarvis Main Brain

Secure admin console

Next version can replace this with username/password sessions. For now the API key protects the admin console.

Dashboard

Live status of Jarvis brain and connected systems.

CheckingSwagger Docs
Brain...FastAPI/Nginx health check.
LLM...Active provider and model.
Workers0Registered worker devices.
Needs Approval0Tasks waiting for admin approval.

Quick command

Ask Jarvis in natural language. Risky actions become approval tasks.

No command run yet.

Recent workers

Recent tasks

AI Chat

Natural-language GPT planner. It chooses safe tools, creates tasks, or queues worker commands.

Jarvis reply will appear here.

Workers

Devices or agents paired to this Jarvis brain.

Send command to worker


        

      


      

        

          
Tasks & Approvals
Review, approve, reject, cancel, and track planned work.

          

          
          
        

        

      


      

        
Agents
Specialist roles GPT can use later.

          

          
        

      


      

        
LLM Settings
Configure OpenAI/GPT or local model. Saved key is masked after saving.

          

          
          

          

        

      


      

        

          
Security checklist
  • Rotate Jarvis API key after screenshots or sharing.
  • Rotate OpenAI key if exposed.
  • Keep Redis/Postgres private inside Docker.
  • Use approval for permission level 3 and 4 tasks.
  • Delete expired pending workers regularly.

          
Audit logs
Recent system actions.