Latest news with #ClaudeCode
Geeky Gadgets
a day ago
- Geeky Gadgets
Claude Code Sub-Agents : The Secret to Smarter Coding Workflows
What if you could delegate the most tedious parts of your coding workflow to specialized assistants that never lose focus, make fewer mistakes, and adapt seamlessly to your needs? Enter Claude Code sub-agents, a new innovation that's reshaping how developers approach complex projects. Imagine debugging, testing, or generating documentation without the mental overhead of juggling multiple tasks. These sub-agents act as purpose-built collaborators, designed to handle specific responsibilities with precision and efficiency. By automating repetitive tasks and maintaining task-specific context, sub-agents free you to focus on what truly matters: solving strategic problems and building great software. In this exploration of the Claude Code Sub-Agents Workflow, World of AI uncover how these modular tools can transform your development process. From reducing errors to improving multitasking, sub-agents are more than just a productivity boost—they're a smarter way to manage complexity. You'll learn how they operate, how to set them up, and the real-world scenarios where they shine, such as debugging or streamlining large-scale projects. Whether you're a solo developer or part of a collaborative team, these insights will reveal how sub-agents can help you work smarter, not harder. As you read on, consider how your current workflow could evolve with tools designed to adapt to your unique challenges. What Are Sub-Agents? Sub-agents are modular, purpose-built tools designed to tackle specific tasks such as debugging, testing, or generating documentation. Each sub-agent operates with its own system prompt, toolset, and context window, making sure precision and minimizing risks such as hallucination or misinterpretation. This targeted approach allows you to delegate tasks effectively, making sure that each sub-agent focuses solely on its area of expertise. By isolating responsibilities, sub-agents enhance accuracy and reduce the likelihood of errors, making them an indispensable part of modern development workflows. How Sub-Agents Work Claude Code functions as a virtual project manager, intelligently assigning tasks to the most suitable sub-agents. These agents are reusable across different projects and teams, offering both flexibility and scalability. With configurable permissions and behavior, you can tailor each sub-agent to meet the unique demands of your workflow. For example, you might configure one sub-agent to handle debugging while another focuses on documentation. This adaptability ensures seamless alignment with your project goals, whether you're working on a small feature or a large-scale application. Claude Code Sub-Agents Workflow Watch this video on YouTube. Gain further expertise in Claude Code by checking out these recommendations. Setting Up Sub-Agents To begin using sub-agents, you need to install Claude Code and ensure you have 18+ installed. The setup process involves defining agent types, tools, and prompts to customize their behavior. This configuration ensures that each sub-agent is equipped to handle specific tasks effectively. Key steps in the setup process include: Identifying the tasks you want to automate, such as debugging or testing. Configuring system prompts to guide each sub-agent's behavior. Assigning the appropriate tools and permissions to each agent. By tailoring sub-agents to your project requirements, you can integrate them seamlessly into your workflow. This not only enhances productivity but also reduces the manual effort required for repetitive tasks. Key Use Cases for Sub-Agents Sub-agents are versatile tools that can be applied to a wide range of development scenarios. Some of the most impactful use cases include: Debugging: Automating the identification and resolution of code errors to save time and improve accuracy. Automating the identification and resolution of code errors to save time and improve accuracy. Documentation: Generating detailed, project-specific documentation to enhance team collaboration and knowledge sharing. Generating detailed, project-specific documentation to enhance team collaboration and knowledge sharing. Testing: Running comprehensive test cases to ensure code reliability and performance. Running comprehensive test cases to ensure code reliability and performance. Task Delegation: Breaking down complex projects into manageable components, allowing better organization and faster completion. These use cases demonstrate the practical value of sub-agents in simplifying workflows and improving overall efficiency. Benefits of Sub-Agents Integrating sub-agents into your coding workflow offers several distinct advantages. These include: Increased Efficiency: By automating repetitive tasks, sub-agents free up your time for higher-level problem-solving and innovation. By automating repetitive tasks, sub-agents free up your time for higher-level problem-solving and innovation. Enhanced Context Management: Each sub-agent maintains task-specific context, reducing errors and confusion during development. Each sub-agent maintains task-specific context, reducing errors and confusion during development. Customizability: You can tailor sub-agents with specific tools and prompts to suit the unique needs of your project. You can tailor sub-agents with specific tools and prompts to suit the unique needs of your project. Scalability: Sub-agents can be reused across multiple projects, making them a cost-effective solution for teams of all sizes. These benefits collectively reduce cognitive load, allowing you to focus on strategic decisions while sub-agents handle operational details. Customization and Flexibility One of the most compelling features of sub-agents is their adaptability. You can customize them at both the project and user levels, making sure they align with your specific requirements. Examples of customization include: Configuring an agent to automate testing scenarios, making sure thorough coverage and reliability. Setting up an agent to streamline the design of user interfaces, improving usability and user experience. Adjusting permissions and workflows to align with team-specific requirements, making sure seamless collaboration. This flexibility allows sub-agents to evolve alongside your projects, adapting to new challenges and priorities as they arise. Whether you're working on a small feature or managing a large-scale application, sub-agents provide the tools you need to stay agile and efficient. Real-World Applications The practical applications of sub-agents are extensive, making them a valuable asset in various development scenarios. For instance, you can use sub-agents to: Automate repetitive tasks in signup workflows, saving time and effort. Design intuitive user interfaces that enhance the overall user experience. Delegate tasks in large-scale projects, making sure timely completion and better organization. Streamline debugging processes, quickly identifying and resolving issues to maintain project timelines. By integrating sub-agents into your projects, you can achieve greater precision and efficiency. This not only improves the quality of your work but also accelerates development timelines, allowing you to deliver better results faster. The Future of Coding Workflows Claude Code sub-agents represent a fantastic step forward in how developers approach coding workflows. By allowing task-specific agents, you can delegate responsibilities, preserve context, and optimize processes with ease. Their adaptability, precision, and efficiency make them a powerful addition to any developer's toolkit. Whether you're tackling a small feature or managing a complex, multi-team project, sub-agents provide the tools you need to simplify workflows and achieve better results. As development challenges continue to evolve, sub-agents offer a scalable and intelligent solution to meet the demands of modern software engineering. Media Credit: WorldofAI Filed Under: AI, Top News Latest Geeky Gadgets Deals Disclosure: Some of our articles include affiliate links. If you buy something through one of these links, Geeky Gadgets may earn an affiliate commission. Learn about our Disclosure Policy.
WIRED
2 days ago
- Business
- WIRED
I Watched AI Agents Try to Hack My Vibe-Coded Websit
Jul 30, 2025 2:00 PM RunSybil, a startup founded by OpenAI's first security researcher, deploys agents that probe websites for vulnerabilities—part of a new AI era for cybersecurity. Illustration:A few weeks ago, I watched a small team of artificial intelligence agents spend roughly 10 minutes trying to hack into my brand new vibe-coded website. The AI agents, developed by startup RunSybil, worked together to probe my poor site to identify weak spots. An orchestrator agent, called Sybil, oversees several more specialized agents all powered by a combination of custom language models and off-the-shelf APIs. Whereas conventional vulnerability scanners probe for specific known problems, Sybil is able to operate at a higher level, using artificial intuition to figure out weaknesses. It might, for example, work out that a guest user has privileged access—something a regular scanner might miss—and use this to build an attack. Ariel Herbert-Voss, CEO and cofounder of RunSybil, says that increasingly capable AI models are likely to revolutionize both offensive and defensive cybersecurity. 'I would argue that we're definitely on the cusp of a technology explosion in terms of capabilities that both bad and good actors can take advantage of,' Herbert-Voss told me. 'Our mission is to build the next generation of offensive security testing just to help everybody keep up.' The website targeted by Sybil was one I created recently using Claude Code to help me sort through new AI research papers. The site, which I call Arxiv Slurper consists of a backend server that accesses the Arxiv—where most AI research is posted—along with a few other resources, combing through paper abstracts for words like 'novel', 'first', 'surprising' as well as some technical terms I'm interested in. It's a work in progress, but I was impressed with how easy it was to cobble together something potentially useful, even if I had to fix a few bugs and configuration issues by hand. A key problem with this kind of vibe-coded site, however, is that it's hard to know what kinds of security vulnerabilities you may have introduced. So when I spoke to Herbert-Voss about Sybil, I decided to ask if it could test my new site for weaknesses. Thankfully, and only because my site is so incredibly basic, Sybil did not find any vulnerabilities. Herbert-Voss says most vulnerabilities tend to be the result of more complex functionality like forms, plugins, and cryptographic features. We watched as the same agents tried probing a dummy ecommerce website with known vulnerabilities owned by Herbert-Voss. Sybil built a map of the application and how it is accessed, probed for weak spots by manipulating parameters and testing edge cases, and then chained together findings, testing hypotheses, and escalating until it breaks something meaningful. In this case, it did identify ways to hack the site. Unlike a human, Herbert-Voss says Sybil runs thousands of these processes in parallel, doesn't miss details, and doesn't stop. 'The result is something that behaves like a seasoned attacker but operates with machine precision and scale,' he says. 'AI-powered pen testing is a promising direction that can have significant benefits for defending systems,' says Lujo Bauer, a computer scientist at Carnegie Mellon University (CMU) who specializes in AI and computer security. Bauer recently coauthored a study with others from CMU and a researcher from AI company Anthropic that explores the promise of AI penetration testing. The researchers found that the most advanced commercial models could not perform network attacks but developed a system that set high-level objectives like scanning a network or infecting a host, which enabled them to perform penetration tests. Sarah Guo, an investor and founder at investment firm Conviction which is backing RunSybil, says it is rare to find people who understand both AI and cybersecurity. Guo adds that RunSybil promises to make the kind of security assessment that large companies perform periodically more widely available, and on a continuous basis. 'They can do baseline penetration testing with models and tool use continuously,' she says. 'So you'll always have a view of what it really looks like to be under attack.' The techniques being developed by RunSybil may become doubly necessary as attackers develop their own AI strategies. 'We have to assume that attackers are already using AI to their benefit,' says Bauer of CMU. 'So developing pen-testing tools that use it is both responsible and likely necessary to balance the increasing risk of attack. Herbert-Voss seems like a good person to help here, since he was the first security researcher at OpenAI. 'I built all sorts of crazy things like new prototypes of polymorphic malware, spearphishing infrastructure, reverse engineering tools,' Herbert-Voss says. 'I was concerned that we didn't have a solution for when everybody gets access to language models—including the bad guys.' This is an edition of Will Knight's AI Lab newsletter. Read previous newsletters here.
Tom's Guide
3 days ago
- Business
- Tom's Guide
Anthropic is putting a limit on a Claude AI feature because people are using it '24/7'
Anthropic has implemented weekly rate limits on its Claude Code tool because it says programmers are using it too much. According to a series of posts from the company, users have been running Claude Code — the agentic side of one of the best ChatGPT alternatives — "continuously in the background, 24/7." Claude Code is only available through Anthropic's paid-for Pro and Max subscription plans. It sits inside the Claude chatbot and uses natural language to help with coding tasks, from generating the code and debugging it to building out applications with a supplied SDK and integrating with Github. The most expensive Max plan, like other AI providers, runs to $200 a month, so you may think the customers are just trying to get their money's worth. But running the tool continuously adds up to some serious costs for Anthropic which has to manage both compute load and environmental costs to provide the service. Some of the biggest Claude Code fans are running it continuously in the background, 24/ uses are remarkable and we want to enable them. But a few outlying cases are very costly to support. For example, one user consumed tens of thousands in model usage on a $200 28, 2025 The company says affected users will start seeing the new weekly rate limits begin in August 28 and says it estimates they'll "apply to less than 5% of subscribers based on current usage." However, it also notes one user consumed "tens of thousands in model usage" on a $200 plan. Get instant access to breaking news, the hottest reviews, great deals and helpful tips. "We're still exploring the best ways to ensure as broad access as possible to Claude Code," the company said in a follow-up tweet. "If you're a power user who has feedback on how we can support your specific use of Claude Code, we'd love to hear from you." For anyone that does go over the limits, they'll have the option to purchase more usage above and beyond their subscription payments. And if you've not used Claude yourself yet, here's why we think Anthropic's AI chatbot is quietly crushing it when compared to ChatGPT or Google Gemini. Follow Tom's Guide on Google News to get our up-to-date news, how-tos, and reviews in your feeds. Make sure to click the Follow button.
Geeky Gadgets
3 days ago
- Business
- Geeky Gadgets
GPT-5 Lobster and Claude Code Combined : The Future of Smarter Automation?
What if the future of coding, creativity, and automation wasn't just faster—but fundamentally smarter? Enter GPT-5 Lobster and Claude Code, two new AI tools that are reshaping how we think about innovation. From transforming workflows to allowing entirely new creative possibilities, these technologies are more than incremental upgrades—they're seismic shifts in the landscape of artificial intelligence. Imagine an AI that not only generates flawless code but also anticipates your needs, or a system that transforms annotated images into cinematic-quality videos. These aren't distant dreams—they're the reality these tools are delivering today. But how do they stack up against the competition, and what do they mean for the future of work? All About AI explores the innovative capabilities of GPT-5 Lobster and Claude Code, diving into their unique features, real-world applications, and the industries they're poised to disrupt. You'll discover how Lobster excels in natural language processing and contextual understanding, while Claude Code pushes the boundaries of agentic engineering to automate complex processes. Whether you're a developer, content creator, or business leader, these tools offer insights into how AI can amplify your productivity and creativity. As we unpack their potential, you might just find yourself rethinking what's possible in your own workflows. AI Automation Tools Driving Innovation GPT-5 Lobster Model: Setting a New Benchmark in AI The GPT-5 Lobster model is a new advancement in natural language processing, excelling in tasks such as text generation, problem-solving, and contextual understanding. Tested rigorously in the LM Web Arena, Lobster consistently outperforms competitors like Starfish, Nectarine, and Kimmy K2. Its ability to interpret context and intent with precision makes it a versatile tool across diverse industries. Content Creation: Lobster generates highly accurate and contextually relevant outputs, streamlining tasks such as drafting articles, creating marketing copy, or even scripting presentations. Lobster generates highly accurate and contextually relevant outputs, streamlining tasks such as drafting articles, creating marketing copy, or even scripting presentations. Data Analysis: It processes complex datasets efficiently, providing actionable insights that support informed decision-making. It processes complex datasets efficiently, providing actionable insights that support informed decision-making. Customer Interaction: Lobster enhances customer service by delivering conversational, accurate, and human-like responses. For industries requiring precision, such as legal documentation or medical research, Lobster's adaptability allows for fine-tuning to meet specific needs. Its ability to cater to unique challenges ensures that it remains a valuable asset for professionals seeking tailored solutions. Claude Code: Transforming Automation with Agentic Engineering Claude Code is transforming automation by allowing the development of autonomous agents designed for coding, project management, and workflow optimization. These agents simplify complex processes, allowing you to focus on strategic and creative tasks. Front-End and Back-End Automation: Claude Code enables the deployment of agents to handle repetitive coding tasks, manage project timelines, and oversee debugging processes. Claude Code enables the deployment of agents to handle repetitive coding tasks, manage project timelines, and oversee debugging processes. Collaboration: By streamlining communication and task delegation, Claude Code fosters seamless teamwork and enhances productivity. With the support of MCP servers, Claude Code is advancing agentic engineering, creating systems that require minimal human intervention. These adaptive systems learn and evolve over time, making them indispensable in dynamic and fast-paced environments. GPT 5 Lobster & Claude Code Automations Watch this video on YouTube. Check out more relevant guides from our extensive collection on ChatGPT 5 that you might find useful. Veo3 Video Generation: Precision Meets Creativity The Veo3 frame-to-video workflow is a fantastic tool in video production, allowing the conversion of annotated images into high-quality videos. This innovation enables you to bring creative visions to life with unparalleled precision and efficiency. Filmmaking: Experiment with camera angles, lighting, and scene compositions without the need for costly equipment or extensive reshoots. Experiment with camera angles, lighting, and scene compositions without the need for costly equipment or extensive reshoots. Marketing: Quickly create engaging and visually appealing content, saving both time and resources while maintaining high production quality. By eliminating the need for labor-intensive post-production processes, V3 opens up new possibilities for creativity while significantly streamlining video creation workflows. This makes it an invaluable tool for filmmakers, marketers, and content creators alike. AI IDEs and CLI Tools: Boosting Developer Productivity AI-powered Integrated Development Environments (IDEs) and Command-Line Interface (CLI) tools are transforming the way developers write, debug, and optimize code. Tools like Gemini CLI, Warp, and Codeex are designed to enhance efficiency and reduce repetitive tasks, allowing developers to focus on innovation. Real-Time Error Detection: Identify and resolve coding issues as they occur, reducing debugging time and improving code quality. Identify and resolve coding issues as they occur, reducing debugging time and improving code quality. Code Optimization: Receive intelligent suggestions to enhance code performance, readability, and maintainability. Receive intelligent suggestions to enhance code performance, readability, and maintainability. Automated Code Generation: Save time by automating boilerplate coding tasks, allowing you to focus on more complex aspects of development. For instance, Gemini CLI's advanced debugging capabilities make it ideal for tackling intricate projects, while Warp's speed-focused design accelerates command execution. These tools are reshaping the development landscape by making coding more efficient and accessible. Runway's ALAP Model: Innovating Video Editing Runway's ALAP model is redefining video editing by offering tools that enhance both flexibility and creativity. This model allows you to reimagine video content in ways that were previously unattainable, making it a fantastic option for professionals in various industries. Generate New Perspectives: Modify camera angles or transform existing footage to create fresh and dynamic visuals. Modify camera angles or transform existing footage to create fresh and dynamic visuals. Create New Sequences: Build entirely new scenes from existing material, expanding the creative potential of your projects. Whether you're working in advertising, entertainment, or content creation, ALAP provides the tools to elevate your work. Its ability to streamline complex editing tasks ensures that you can focus on delivering high-quality, impactful content. Unitree R1: Humanoid Robotics with LLM Integration The Unitree R1 humanoid robot, integrated with large language models (LLMs), represents a significant leap forward in robotics. Designed for versatility, the R1 is capable of performing tasks across a wide range of applications, from customer service to advanced research. Retail Assistance: Deploy the R1 to guide customers, answer queries, and enhance the overall shopping experience. Deploy the R1 to guide customers, answer queries, and enhance the overall shopping experience. Laboratory Automation: Automate repetitive tasks in research settings, allowing human workers to focus on strategic and analytical roles. With its ability to understand and respond to complex commands, the R1 is adaptable to various industries. Its integration with LLMs ensures that it remains a valuable tool for innovation, efficiency, and problem-solving. The Future of AI in Development AI is no longer just a tool; it is a driving force behind innovation and progress. By integrating AI technologies like GPT-5 Lobster, Claude Code, and advanced video generation tools into your workflows, you can unlock new levels of efficiency, creativity, and adaptability. These advancements are not only reshaping industries but also redefining the way you work and solve problems. The tools highlighted here demonstrate the fantastic potential of AI across multiple domains. By embracing these technologies, you can stay ahead of the curve, using AI to achieve your goals and shape the future of work in a rapidly evolving world. Media Credit: All About AI Filed Under: AI, Top News Latest Geeky Gadgets Deals Disclosure: Some of our articles include affiliate links. If you buy something through one of these links, Geeky Gadgets may earn an affiliate commission. Learn about our Disclosure Policy.
India Today
3 days ago
- Business
- India Today
Claude AI to get new weekly usage limits as Anthropic cracks down on 24x7 use, account sharing
Anthropic has announced new weekly usage restrictions for Claude AI, aimed at preventing people from running its coding tool non-stop or sharing accounts with others. The fresh limits will be introduced from August 28 and will apply to users across all paid plans, including the $20 Pro tier and the higher-priced $100 and $200 Max an email to users and a post on social media, Anthropic said some people were keeping Claude Code running continuously in the background or violating its rules by reselling access or sharing login details. The company now plans to introduce two new weekly caps: one on total usage, and another specifically for the Claude Opus 4 model, its most advanced company clarified that the current usage limits (which refresh every five hours) will remain unchanged. However, users on the Max plans will now have an option to buy extra access once they hit their weekly cap, using standard API pricing. These changes come at a time when Claude Code, the AI coding assistant, has been seeing increased demand among developers. But this growing popularity has also brought some challenges. According to Anthropic's system status page, the tool has faced several outages in the past month, possibly due to some users running it around the clock."Claude Code has experienced unprecedented demand since launch," Anthropic spokesperson Amie Rotherham told TechCrunch in an email. She also said that "most users won't notice a difference," and that the new limits are expected to affect fewer than 5 per cent of per the updated plans, subscribers of the Pro tier can expect 40 to 80 hours of Claude Code using Sonnet 4 each week. Those on the $100 Max plan will get 140 to 280 hours of Sonnet 4 and 15 to 35 hours of Opus 4. Users on the $200 Max plan will be allowed 240 to 480 hours of Sonnet 4 and 24 to 40 hours of Opus 4 in a didn't explain how exactly usage is being tracked — whether it's by number of tokens used, compute time, or hours spent. While the company earlier claimed the $200 plan offers 20 times more access than the Pro tier, the latest numbers suggest the increase may be closer to six times in actual usage move follows a trend among AI tool providers, who are reworking their pricing and usage models to prevent abuse. In June, Anysphere, the team behind Cursor, made similar changes for its Pro plan, which led to confusion and criticism after users were unexpectedly charged more. Around the same time, another company, Replit, also adjusted its pricing structure.- Ends
