Anthropic launched a new capability on Thursday that allows that Claude A Assistants to acquire specialized skills on demand, marking the company’s latest effort to make artificial intelligence more practical for enterprise workflows, as it chases rival Openai in a fiercely competitive battle against AI-powered software development.
feature, called Skillenables users to create folders containing instructions, code scripts, and reference material that Cloud can automatically load when relevant to a task. The system marks a fundamental shift in how organizations can customize AI assistants, moving beyond one-off cues to reusable packages of domain expertise that work consistently across the company.
"Expertise is based on our belief and vision that as model intelligence improves, we will continue to move towards general-purpose agents that frequently access our file systems and computing environments," Mahesh Murg, member of Anthropic’s technical staff, said in an exclusive interview with VentureBeat. "The agent is initially only informed of the names and descriptions of each available skill and may choose to load more information about a particular skill when relevant to the task at hand."
The launch comes like Anthropic, which has a price 3 183 billion after the recent funding round of 13 billion dollarsAccording to a recent report, its annual revenue could nearly triple to $26 billion in 2026. Reuters reports. The company is currently approaching $7 billion in annual revenue, up from $5 billion in August, largely fueled by enterprise adoption of its AI coding tools.
How ‘progressive disclosure’ solves the context window problem
Skill That’s fundamentally different from existing methods for customizing AI assistants, such as rapid engineering or recursive generation (RAG), Morag explained. Architecture relies on what anthropic calls "Progressive disclosure" – Cloud initially only sees skill names and short descriptions, then autonomously decides which skills to load based on the task, while only accessing specific files and information.
"Unlike Rig, it relies on simple tools that let Cloud manage and read files from the file system," Morag told VentureBeat. "Skills can include an infinite amount of context to teach Claude how to complete a task or series of tasks. This is because skills are based on an agent that can autonomously and intelligently execute a file system and code."
This approach allows organizations to bundle more information than traditional contextual Windows permits, while maintaining the speed and performance that enterprise users demand. A single skill can include step-by-step procedures, code templates, reference documents, brand guidelines, compliance checklists, and executable scripts. All of these are organized in a folder structure that Cloud navigates intelligently.
System composability provides another technical advantage. Multiple skills automatically stack together when needed for complex workflows. For example, Claude may simultaneously call for expertise in a company’s brand guidelines, expertise in financial reporting, and expertise in presentation formatting to produce a quarterly investor deck.
Skills that differentiate it from Openai’s Custom GPTS and Microsoft’s Copilot
Anthropic has expertise that sets it apart from competing offerings like Openai Custom GPTS And Microsoft’s Copilot Studiothough the features address similar enterprise needs around AI customization and consistency.
"Skills’ combination of progressive disclosure, composability, and executable code bundling is unique in the market," Death said. "While other platforms require developers to build custom scaffolding, Expertise allows anyone – technical or not – to build specialized agents by organizing procedural knowledge into files."
Cross-platform portability also sets the skill apart. The same skill works equally well Claude Efor , for , for , . Claude Code (Anthropic’s AI coding environment), the company’s apiand Cloud Agent SDK For building custom AI agents. Organizations can build expertise once and deploy it everywhere their teams use the cloud, which is a key benefit for enterprises to achieve consistency.
The feature supports any programming language compatible with the underlying container environment, and provides sandboxing for Anthropic security — although the company acknowledges that allowing AI to execute code requires users to carefully evaluate which skills they trust.
Early users report 8x productivity gains on finance workflows
Early user implementation shows how organizations are applying Skill Automating complex knowledge tasks. In the Japanese e-commerce giant RakutenAI teams are using expertise to transform finance operations that previously required manual coordination across multiple departments.
"Expertise streamlines our management accounting and finance workflows," Yusuke Kaji, general manager of AI at Rakuten, said in a statement. "Claude processes multiple spreadsheets, catches critical anomalies, and generates reports using our methodology. What once took a day, we can now accomplish in an hour."
That’s an 8x improvement in productivity for specific workflows—the kind of measurable return on investment that enterprises increasingly demand from AI implementation. Mike Krieger, chief product officer of Anthropic and co-founder of Instagram, recently noted that companies have moved past "Ai Fomo" Concrete success measurements are needed.
Design platform Canva It plans to integrate the skill into its AI agent workflows. "Canva aims to leverage skills to customize agents and extend what they can do," Anwar Hanif, general manager and head of ecosystem at Canva, said in a statement. "This opens up new ways to bring Canva deeper into agent workflows."
Cloud storage provider Box Sees specialization as a way to make corporate content repositories more actionable. "Skill teaches Claude how to work with the contents of the box," said Yashodha Bhunani, head of AI in the box. "Users can convert stored files into PowerPoint presentations, Excel spreadsheets, and Word documents that adhere to their organization’s standards."
Enterprise Security Question: Who controls who can use which AI skills?
For enterprise IT departments, Skill Raise important questions about governance and control – especially since this feature allows AI to execute arbitrary code in a sandboxed environment. Anthropic has built in administrative controls that allow enterprise users to manage access at an organizational level.
"Enterprise admins control access to skill capabilities through admin settings, where they can enable or disable access and monitor usage patterns," Death said. "Once enabled at the organizational level, individual users still need to opt-in."
This two-layer consent model—organizational capability plus individual opt-in—reflects lessons learned from previous enterprise AI deployments where blanket rollouts have raised compliance concerns. However, Anthropic’s governance tools appear more limited than some enterprise users expect. The company does not currently offer granular control over which specific skills employees can use, or detailed audit trails of custom skill content.
Organizations concerned about data security should note that the skills require a cloud code execution environment, which runs in isolated containers. Anthropic advises consumers "Stick to reliable sources" While installing the skill and providing security documentation, the company acknowledges that this is an inherently higher risk potential than traditional AI interaction.
From API to no-code: How Anthropic is making expertise accessible to everyone
Anthropic is taking multiple approaches to create Skill Accessible to users of varying technical sophistication. For non-technical users Claude Ethe company provides a "A skill builder" A skill that interactively guides users through building new skills by asking them questions about their workflow, then automatically generates folder structures and documents.
Working with developers Anthropic’s API Gain programmatic control through a new /skill endpoint and manage skill versions through the cloud console web interface. API requests for the feature require the code execution tool beta to be enabled. For CloudCode users, skills can be installed through the Anthropics/Skills GitHub Market plugin, and teams can share skills through a version control system.
"Expertise is included in Max, Pro, Teams and Enterprise plans at no additional cost," Murg confirmed. "API usage determines standard API pricing," Meaning organizations only pay for the tokens used during skill execution, not for the skill itself.
Anthropic provides several pre-built skills for common business tasks, including professional generation Excel spreadsheets with formulas, PowerPoint presentations, Word documents, and fillable PDFs. These anthropogenically generated skills will remain free.
Why Skills Launch Matters in the AI Coding Wars with OpenAI
The talent announcement comes during an important moment for Anthropic with OpenAI, particularly around AI-assisted software development. Just a day before releasing the skill, Anthropic launched Claude Haikou 4.5a smaller and cheaper model that nevertheless matches the coding performance Claude Sonnet 4 -which was state-of-the-art when it was released just five months ago.
This steep improvement curve reflects the rapid pace of AI development, where today’s frontier capabilities become tomorrow’s commodity offerings. OpenAI is also putting a strong emphasis on coding tools, recently upgrading it Codex Platform with the GPT-5 And spreading of GitHub Coplot Abilities
Anthropic’s revenue path – potentially reachable 26 billion in 2026 An estimated $9 billion by year-end 2025—suggests the company is successfully converting enterprise interest into paying customers. The timing also follows Salesforce’s announcement this week that it has deepened AI partnerships with both OpenAI and Anthropic to power its Agent Force platform, signaling that enterprises are embracing a multi-vendor approach rather than standardizing on a single provider.
Skills identify a real pain point: "Quick Engineering" The problem is that where effective AI use depends on individual employees developing broad instructions for routine tasks, there’s no way to share that expertise across teams. Skills transform tangible knowledge into tangible, shareable assets. For startups and developers, the feature can significantly speed up product development — including advanced document generation capabilities that previously required dedicated engineering teams and weeks of development.
The composability aspect points to a future where organizations build libraries of specialized skills that can be mixed and matched for increasingly complex workflows. A pharmaceutical company can develop expertise for regulatory compliance, clinical trial analysis, molecular modeling, and patient data privacy that work seamlessly together.
Anthropic indicates that it is working on simplified talent creation workflows and enterprise-wide deployment capabilities to make it easier for organizations to distribute skills across large teams. As this feature is offered to Anthropic’s more than 300,000 business customers, the real test will be whether organizations can leverage existing customization methods beyond that.
For now, the skills have yet to provide Anthropic’s clearest explanation of its vision for AI agents: not generalists who try to do every task reasonably well, but intelligent systems that know when to access specialized expertise and can integrate multiple domains of knowledge to accomplish complex tasks. If this vision plays out, the question won’t be whether your company uses AI — it will be whether your AI knows how your company actually works.