Raft 1.0 发布：多智能体长期协作平台

There are more and more multi-agent AI systems everywhere, but still, every handoff needs someone to keep everything together. Raft just launched to turn them into long-term collaborators with their own memories, roles, shared threads, and the ability to review each other’s work. Every agent has its own identity, memory and role that it plays in every project. They are separate and thus an agent can challenge another without grading his own work. Each thread logs decisions, files, reviews and open tasks so work can continue in parallel. People still decide what to do first, look at results and make final decisions before anything else happens. Raft claims it has more than 100 named agents now doing 99% of its internal work. The beta had over 20,000 builders, or roughly 4 agents per human.     > **引用原帖 stdrc (@istdrc):** > Hi, I'm RC. I built Kimi CLI at Moonshot last year, and back in 2015, bots that lived in group chats. For the past four months, I've been building Raft in public. > Today I'm launching Raft 1.0. > Right now, working with agents means juggling terminals, sessions, and skills. The more you run, the more you end up holding it all together yourself, and the easier it is to lose the thread. > Raft puts your agents in team mode: one workspace where working with agents feels like messaging your team. The work keeps moving, and you stay at the wheel. > Meet my Raft agent team👇 > https://x.com/istdrc/status/2077376131628707907