夜雨聆风
Since the beginning of 2026, an unprecedented "shrimp-raising" craze has swept across the global tech and financial circles, driven by the popularity of OpenClaw, an open-source AI agent with a red lobster icon. This craze refers to the practice of deploying and training OpenClaw to complete various tasks automatically, which has been favored by securities analysts, office workers and tech enthusiasts alike.
2026年初，随着带有红色龙虾图标的开源AI智能体OpenCLAW的走红，一场前所未有的“养虾”热潮席卷了全球科技圈和金融圈。这里的“养虾”指的是部署并训练OpenCLAW，让其自动完成各类任务，这种方式已受到证券分析师、职场人士和科技爱好者的广泛青睐。
OpenClaw, formerly known as Clawdbot and Moltbot, is a local-first AI agent framework that can take over users’ keyboard and mouse permissions, operate computer terminals independently, write code, manage files, and even learn new "skills" according to users’ natural language instructions. Unlike cloud-based large models that only provide text solutions, it runs entirely on local devices or private clouds, making it more flexible and private.
OpenCLAW 前身为Clawdbot和Moltbot，是一款本地优先的AI智能体框架，能够接管用户的键鼠权限，自主操作电脑终端、编写代码、管理文件，甚至能根据用户的自然语言指令学习新的“技能”。与仅提供文本解决方案的云端大模型不同，它完全运行于本地设备或私有云，灵活性和私密性更强。
In the financial research and investment industry, OpenClaw has become a "super employee" that liberates analysts’ hands. Many securities firms, including Founder Securities and GF Securities, have released special reports to teach practitioners how to deploy OpenClaw, covering scenarios such as automated information processing, research report writing, and quantitative strategy development.
在金融投研行业，OpenCLAW 已成为解放分析师双手的“超级员工”。包括方正证券、广发证券在内的多家券商，已发布专题报告，教从业者如何部署OpenCLAW，涵盖自动化信息处理、研报撰写、量化策略开发等场景。
While the "shrimp-raising" craze has brought remarkable efficiency improvements—helping users save time on repetitive tasks and improve work efficiency—it has also raised concerns about potential risks. Security risks such as file deletion or information leakage may occur if users lack security awareness during configuration, and AI hallucinations may lead to incorrect data and conclusions.
尽管“养虾”热潮带来了显著的效率提升——帮助用户节省重复任务的时间、提高工作效率——但也引发了人们对潜在风险的担忧。如果用户在配置过程中缺乏安全意识，可能会出现文件删除、信息泄露等安全风险，而AI幻觉则可能导致数据和结论出错。
To address these risks, experts suggest that users should deploy OpenClaw in an isolated environment separate from work or personal computers. In addition, the OpenClaw development team plans to optimize the system continuously, improve security mechanisms, and standardize the ecological environment of third-party "skills" to promote the healthy development of the "shrimp-raising" craze.
为应对这些风险，专家建议用户应将OpenCLAW部署在与工作或个人电脑隔离的环境中。此外，OpenCLAW开发团队计划持续优化系统、完善安全机制、规范第三方“技能”的生态环境，以推动“养虾”热潮的健康发展。
