Anthropic Launches "Skills" to Boost Claude’s Workplace AI Capabilities
Key Points
- Anthropic launches "Skills" for Claude, a set of task‑specific instructions and scripts.
- Skills aim to reduce prompt‑engineering effort and improve task accuracy.
- Box, Rakuten and Canva have already tested the feature.
- Users can create custom Skills for their own workflows across multiple Claude platforms.
- OpenAI recently unveiled AgentKit, a competing suite for building production‑ready AI agents.
- Albertsons used an OpenAI agent to address a 30% drop in ice‑cream sales across its 2,000 U.S. stores.
- Both companies are pushing AI agents toward practical enterprise use.
Anthropic announced a new feature called Skills for Claude, a collection of pre‑built instructions, scripts and resources that let the Claude AI agent perform specific work tasks more efficiently. Companies such as Box, Rakuten and Canva have already used the tool, and users can also create custom Skills for their own workflows across Claude.ai, Claude Code, the API and the Claude Agent SDK. The rollout aims to reduce prompt‑engineering effort and make AI agents more reliable in real‑world business contexts, positioning Anthropic alongside rivals like OpenAI, which recently unveiled its own AgentKit suite.
Anthropic Introduces Skills for Claude
Anthropic unveiled a new capability for its Claude AI model called Skills. The feature bundles together folders that contain instructions, scripts and other resources that Claude can load on demand to handle specific work‑related tasks. According to the company’s release, Skills cover functions ranging from spreadsheet manipulation to adhering to an organization’s brand guidelines. The tool is accessible to Pro, Max, Team and Enterprise users and can be employed across Claude.ai, Claude Code, Anthropic’s API and the Claude Agent SDK.
How Skills Enhance AI Agent Productivity
The primary goal of Skills is to lessen the need for users to craft perfect prompts or repeatedly supply contextual information. By loading a relevant Skill, Claude can execute a task with higher accuracy and speed, effectively acting as a more specialized agent. Anthropic’s product lead Brad Abrams described Skills as a way for organizations to teach Claude to perform well in their specific context, emphasizing practical outcomes over abstract benchmarks.
Early Adoption by Enterprises
Box, Rakuten and Canva are among the first companies to experiment with Skills, according to the announcement. These early adopters have used the feature to streamline internal processes, such as generating presentations or ensuring marketing materials follow brand standards. Users also have the option to build custom Skills tailored to their unique workflows, extending the utility of Claude across a broad range of professional scenarios.
Competitive Landscape and Parallel Developments
Anthropic’s move follows similar initiatives from other AI leaders. Earlier this month, OpenAI introduced AgentKit, a suite of tools designed to help developers move AI agents from prototype to production. OpenAI highlighted a case where Albertsons, a grocery chain operating more than 2,000 U.S. stores, used a custom agent to devise a plan for boosting ice‑cream sales after a decline of over 30 percent. Companies such as Box, Canva, Evernote and Ramp were also mentioned as participants in OpenAI’s testing.
Implications for the Future of AI Agents
Both Anthropic and OpenAI are focusing on making AI agents genuinely useful in day‑to‑day work environments. By providing modular, task‑specific capabilities, they aim to bridge the gap between experimental AI tools and reliable enterprise solutions. The introduction of Skills suggests a shift toward more plug‑and‑play AI functionalities that can be quickly integrated into existing business processes without extensive custom development.