有了Agentic Studio, LinearB提供了加速软件交付所需的关键背景
应用开发智能启动 LinearB Inc. 称今天它正在试图缓解随着人工智能编码工具的广泛采用而出现的组织软件输送管道日益拥塞的现象。
它与代理工作室(Agentic Studio)合作
启动的称出现基因化AI编码工具如Claude Code和Codex导致企业软件开发团队出现新的头痛。 虽然现在写软件代码基本上可以无功而返，但是将代码运入生产过程仍然像往常一样痛苦，现在他们还有更多的代码可以运出。
在其最近的软件工程基准测试报告中，公司分析了250个组织超过270万个拉动请求，发现那些依赖AI编码工具的人已经几乎比年年翻了一番合并代码变化。 但问题是AI协助的拉力请求近2.6倍，AI agent提议修改的代码在“精英组织”中只有79%的时间被合并，而人类授权的代码只有92%。
因为传统的仪表板只能提供衡量标准，在Agentic Studio中，它试图通过将Git寄存器和项目跟踪工具生成的数据和见解正常化，来解决这个问题。 意思是，软件工程师可以不用通过各种图表来尝试和理解释放速度为何停滞的原因，而是用简单的语言提问，以了解根本原因，并接受关于如何加快速度的建议。
共同创始人兼首席执行官Ori Keren 称背景成为新国王，不仅为AI工具，也为人类。 称表示：「我们意识到，agents和团队需要更快地提供高质量的软件
Agentic Studio的不同之处在于它的设计帮助自主编码agents，就像人类软件工程师Keren 称一样。 通过"模型背景协议"等协议的整合，它为AI agents提供了重新工作率，过去生产事件和代码所有权等历史背景，防止他们盲目修改以拉动请求。
早期的收养者从Agentic Studio的生产力大幅提高中受益。 例如密码货币交换(Kraken)，它发现时区错位是主要罪魁祸首，而不是拉动请求大小，随后重组了工作流程，在短短4个月内将审查时间缩短一半以上。
称 LinearB的首席操作官Dan Lines。 “一个仪表板称，你的周期是两天，但又如何？现在[和Agentic Studio一起]，一个领导者可以问他们的工程组织有什么改进，得到一个明确的答案。 当同一个背景去编码agents时
为了帮助AI agents操作它所提供的上下文，LinearB正在开发一批新的agent技能。 第一个是“努力技能”，现在可以及早访问，并分析牵引请求的复杂性及其拟议的代码变化、其重制历史和对编码agent在操作时需要花费多少计算努力进行动态校准的风险。
简单，低风险的更改用最小的符号进行得更快。 将进行更敏感、更优先的修改，进行更深入的推理和更详尽的测试，以确保不会出错。
计划在未来几个月扩充agent技能库
图像：线性B
SiliconANGLE联合创始人John Furrier的留言：
支持我们的使命，通过与CUBE社区接触，保持内容的开放和自由。 加入CUBE的校友信托网络，其中技术领导者相互连接，分享情报，创造机会。
- 15M+观看CUBE视频，为AI、云层、网络安全等各种对话提供动力
- 11.4k+CUBE校友——通过独特的信任网络与11400多名技术和企业领袖建立联系，塑造未来
你是AWS客户吗？在财务上支持SiliconANGLE，从我们的市场门户页面和链接购买您的AWS服务：https://siliconangle.com/aws-marketplace/.
关于 SiliconANGLE 媒体
SiliconANGLE Media由技术视觉家约翰·弗瑞尔(John Furrier)和戴夫·韦兰特(Dave Vellante)创立，构建了产业领先数字媒体品牌的动态生态系统，达到1500+百万精英技术专业人士。 利用CUBEAi.com神经网络，帮助科技公司以数据为导向做出决策
本文由 AI图鉴自动抓取并翻译，版权归原作者所有；如需查看原文请点下方链接。
With Agentic Studio, LinearB provides the critical context needed to accelerate software delivery
Application development intelligence startup LinearB Inc. said today it’s trying to ease the growing congestion in organization’s software delivery pipelines that has arisen with the widespread adoption of artificial intelligence coding tools.
It’s doing so with Agentic Studio, a new offering announced today that’s focused on providing more actionable context for human engineers and autonomous coding agents alike.
The startup says the emergence of generative AI coding tools such as Claude Code and Codex has led to new headaches for enterprise software development teams. Though writing software code now can be basically effortless, shipping that code into production is still as painful as it has always been, and now they have much more of it to ship.
In its recent Software Engineering Benchmarks report, the company analyzed more than 2.7 million pull requests across 250 organizations and found that those who rely on AI coding tools have almost doubled their merged code changes year-over-year. But the problem is that AI-assisted pull requests are nearly 2.6-times larger, and consequently they’re waiting more than five-times longer to be reviewed. This has led to significant backlogs and it’s compounded by the fact that AI agent’s proposed code changes are merged just 79% of the time at “elite organizations,” compared with 92% for human-authored code.
LinearB says that human software engineers struggle to clear this backlog because the traditional dashboards they’re forced to work with only provide metrics, rather than the context needed to resolve those pull requests. With Agentic Studio, it’s trying to address this by normalizing the data and insights generated by Git repositories and project tracking tools into a unified context layer. It means that instead of muddling through various charts to try and understand why release velocity has stalled, software engineers can ask questions in plain language to understand the root cause and receive recommendations on how to speed things up.
Co-founder and Chief Executive Ori Keren said context has become the new king, not only for AI tools but also for humans. “We realized we hold a knowledge graph that nobody else has of everything that is happening across the software delivery lifecycle,” he said. “That’s exactly what agents and teams need to deliver high quality software faster, so we rebuilt LinearB around it.”
What’s different about Agentic Studio is that it’s designed to help autonomous coding agents just as much as human software engineers, Keren said. By integrating through protocols such as the Model Context Protocol, it feeds historical context such as rework rates, past production incidents and code ownership to AI agents, preventing them from making blind modifications to pull requests.
Early adopters have benefited from dramatically increased productivity with Agentic Studio. A case in point is the cryptocurrency exchange Kraken, which used LinearB’s MCP integration to pinpoint why its code reviews were lagging. It discovered that timezone misalignments were the main culprit rather than pull request sizes, and subsequently reorganized its workflows to cut its review times by more than half within just four months.
“For years, engineering leaders told us the same thing: metrics are useful, but they don’t tell you what to do,” said LinearB’s chief operating officer Dan Lines. “A dashboard says your cycle time is two days, but so what? Now [with Agentic Studio] a leader can ask what to improve in their engineering organization and get a clear answer. When that same context goes to coding agents, the platform is not just reporting, it’s acting.”
To help AI agents operationalize the context it provides, LinearB is developing a bunch of new agent skills. The first one, an “effort skill,” is available in early access now and analyzes the complexity of a pull request and its proposed code changes, its rework history and risk to calibrate dynamically how much computational effort a coding agent needs to expend as it works on it.
Simple, low-risk changes are done faster with minimal tokens. More sensitive, high-priority modifications will be implemented with deeper reasoning and more exhaustive testing to ensure nothing goes wrong.
LinearB said it’s planning to expand its agent skills library in the coming months with additional skills focused on automated code review, merge governance and token cost optimization.
Image: LinearB
A message from John Furrier, co-founder of SiliconANGLE:
Support our mission to keep content open and free by engaging with theCUBE community. Join theCUBE’s Alumni Trust Network, where technology leaders connect, share intelligence and create opportunities.
- 15M+ viewers of theCUBE videos, powering conversations across AI, cloud, cybersecurity and more
- 11.4k+ theCUBE alumni — Connect with more than 11,400 tech and business leaders shaping the future through a unique trusted-based network
Are you an AWS customer? Support SiliconANGLE financially by buying your AWS services from our Marketplace portal page and links: https://siliconangle.com/aws-marketplace/
About SiliconANGLE Media
Founded by tech visionaries John Furrier and Dave Vellante, SiliconANGLE Media has built a dynamic ecosystem of industry-leading digital media brands that reach 15+ million elite tech professionals. Our new proprietary theCUBE AI Video Cloud is breaking ground in audience interaction, leveraging theCUBEai.com neural network to help technology companies make data-driven decisions and stay at the forefront of industry conversations.
Auto-fetched and translated by AI Atlas. All rights belong to the original author.