Skyfall AI 用智能体团队 Steve 接管 3 家被收购公司运营
Skyfall AI 买了 3 家线上公司，让智能体团队 Steve 直接管客服、营销和改代码，第一家 MRR 三个月涨了 2.3 倍。和一般 agent demo 不同，它有真实营收数据和'一次收购的经验复用到下一次'的机制。做智能体产品的人可以看看 harness 怎么设计。
Skyfall AI 收购了 3 家线上业务，并让一套名为 Steve 的 AI 智能体团队负责客服、营销和代码修改，人类只在重大决策上把关。第一个被收购的社交媒体应用在不到 3 个月里 MRR 增长 2.3 倍。
Steve 在每个复盘周期记录所学经验并带入下一轮运营，Skyfall 表示第一笔收购的教训直接形成了第二笔收购的运营手册。第二个收购对象是一个 Shopify 应用，在 Steve 重写商品列表、砍掉只带来点击不带购买量的广告后，业务已经企稳。
技术上，两家公司都运行在一个代码世界模型上，每家业务对应一个可执行程序，运营动作上线前先在程序里推演。这个案例也印证了一个趋势：当前最有价值的智能体工程，正越来越多发生在模型外层的 harness 设计上，而不是模型本身。
Another cool launch showing how the most interesting agent engineering right now is happening in the harness around the model.
Skyfall AI bought 3 online businesses and put Steve, a team of AI agents, in charge of support, marketing and code changes, while a human owner approves only the big calls.
Steve records what it learns in every review cycle and carries it forward, and Skyfall says lessons from its first acquired business shaped the playbook for the second.
The 1st, a social media app, grew MRR 2.3x in under 3 months, and the 2nd, a Shopify app, has stabilized after Steve rewrote its listing and cut ads that drove clicks without purchases. Both run on a code world model, one executable program per business, where moves get rehearsed before they go live.
This report is categorized under AITOP's "Products" desk, focusing on US frontier AI labs and Silicon Valley technology leaders. Within our cross-regional analytical framework, it directly maps to AI developer toolchains, coding agents, and engineering workflows and US-China foundation model benchmarks and commercialization comparison. Key entities and technical topics under continuous tracking include Steve, Skyfall AI, 智能体, Shopify.
Source attribution: curated from rohanpaul_ai (@rohanpaul_ai).
