500 bugs let the AI fix itself, and it still has to write weekly reflections. This isn't a tool—it's clearly like hiring an intern who can work 24/7 without paying social insurance. The double confirmation mechanism is just to keep this intern from getting too cocky.

CoinWorld News, Anthropic recently disclosed the engineering experience of human-machine collaborative teams. Multiple employees work together with multiple agents that have independent system credentials in Slack. The agents are directly mounted under the team roster and communication threads, with clear division of labor and autonomous project advancement, just like human employees. To effectively integrate agents into the team, collaboration defaults to making work public. The company sets security boundaries to avoid tedious single-document authorization decisions. The team assigns specialized roles to agents by writing skill files to prevent information fragmentation. The autonomy of agents is proportional to their reliability. Engineering managers dispatched agents to independently fix 500 bugs and required them to submit weekly reflection reports to avoid repeating mistakes. The team uses a double confirmation mechanism, where one agent reviews the work of another, ensuring the sustainable operation of the human-machine team.