Architecture overview for using StateSet Computer Use Agents
main.py
)main()
: Entry point that initializes and runs agentsget_active_agents()
: Determines which agents to activate based on instruction keywordsrun_agent()
: Executes a single agent with error handling and billingcontinuous_loop()
: Manages multiple agents running in parallelanalyze_task_completion()
: Determines if a task was completed successfullyagent/loop.py
)sampling_loop()
: Main conversation loop with the AI modelinitialize_system_prompt()
: Fetches agent configuration from APIs