Anthropic announced Thursday that it will release its Agent Skills framework as an open standard, marking a strategic move to establish industry wide specifications for teaching AI assistants specialized tasks.
The San Francisco based company also unveiled enterprise management tools and a directory of partner built skills from major software providers.
The specification and reference SDK are now available at agentskills.io, making the technology accessible to developers building on any compatible platform.
What Are Agent Skills?
Agent Skills represent a new approach to making AI assistants more capable at specific tasks. Rather than relying on generic prompts, the framework packages instructions, scripts, and resources into organized folders that AI models can load dynamically when needed.
Think of it like creating an onboarding guide for a new employee. Instead of explaining everything from scratch each time, you hand them a structured manual with step-by-step procedures. The AI loads only relevant information when tackling a particular task, keeping its working memory efficient.
“We believe AI skills should not be locked to a single platform,” said Mahesh Murag, product manager at Anthropic, in an interview with VentureBeat. “Developers can write skills that work seamlessly across Claude and other AI platforms that adopt the standard.”
Industry Adoption Already Underway
The open standard has attracted immediate interest from major players. Microsoft has integrated Agent Skills within VS Code and GitHub. Popular coding assistants including Cursor, Goose, Amp, and OpenCode have also adopted the specification.
Perhaps most notably, OpenAI has quietly implemented structurally identical architecture in both ChatGPT and its Codex CLI tool. Developer Elias Judin discovered the implementation earlier this month, finding directories containing skill files that mirror Anthropic’s specification the same file naming conventions, metadata format, and directory organization.
This convergence suggests the industry has found common ground on a challenging question: how to make AI assistants consistently good at specialized work without expensive model fine-tuning.
Enterprise Features for Scale
The update introduces organization wide management capabilities for Team and Enterprise plan subscribers. Administrators can now provision skills centrally, controlling which workflows are available across their organizations while allowing individual employees to customize their experience.
“Enterprise customers are using skills in production across both coding workflows and business functions like legal, finance, accounting, and data science,” Murag explained.
Anthropic has partnered with ten major software companies for the launch. Atlassian brings skills for Jira and Trello task management. Figma offers brand style guideline enforcement. Notion, Stripe, Canva, Cloudflare, and Zapier round out the initial partner roster.
Building on MCP Success
The move follows Anthropic’s success with its Model Context Protocol, which became the de facto communication standard for AI agents. Earlier this month, Anthropic donated MCP to the Linux Foundation, and both Anthropic and OpenAI co-founded the Agentic AI Foundation alongside Block. Google, Microsoft, and Amazon Web Services joined as members.
“We have seen how complementary skills and MCP servers are,” Murag noted. “MCP provides secure connectivity to external software and data, while skills provide the procedural knowledge for using those tools effectively.”
The Agent Skills GitHub repository has already crossed 20,000 stars, with tens of thousands of community-created skills shared across the platform.
Internal Research Validates Approach
Anthropic’s decision stems partly from internal findings. A study published in early December found that company engineers used Claude in 60% of their work, achieving a 50% self-reported productivity boost – a two to threefold increase from the prior year.
Notably, 27% of Claude-assisted work consisted of tasks that would not have been done otherwise, including building internal tools, creating documentation, and addressing small quality-of-life improvements that had been perpetually deprioritized.
What This Means for the Industry
The release signals a broader shift in the AI market. Competition is moving beyond model capabilities toward infrastructure and standards. Companies that define how AI assistants work may capture more value than those trying to own proprietary technology outright.
“The industry is now moving beyond the models into what can the models do on my behalf,” said Arun Chandrasekaran, analyst at Gartner. “How can the models orchestrate workflows? How can models re-engineer processes? How can AI become more autonomous within our environment?”
For enterprise technology leaders evaluating AI investments, the message is straightforward: skills are becoming infrastructure. The expertise organizations encode into skills today will determine how effectively their AI assistants perform tomorrow, regardless of which model powers them.
Sources:
