📊 Full opportunity report: A Skill Is a Folder, Not a Prompt: What Anthropic Learned Running Hundreds of Them on ThorstenMeyerAI.com — validation score, market gap, and execution plan.
TL;DR
Anthropic has demonstrated that modeling AI skills as folders containing instructions and assets, rather than simple prompts, improves consistency, onboarding, and organizational knowledge. This approach is based on running hundreds of such skills internally.
Anthropic has revealed that its internal AI capabilities are built around ‘Skills’ structured as folders containing instructions, scripts, and assets, rather than simple prompts. This shift aims to make AI output more consistent, improve onboarding, and institutionalize organizational knowledge, marking a significant evolution in AI engineering practices.
In a detailed write-up from a Claude Code engineer, Anthropic explained that a Skill is not just a saved prompt but a folder that can include instructions, reference documents, scripts, templates, data, and configuration. The agent can discover and execute these components dynamically, enabling more durable and reusable capabilities.
Anthropic’s internal experience shows that organizing skills this way allows for better standardization of output, easier onboarding of new team members, and the ability to improve skills over time as they are refined with new edge cases. The company emphasizes that these skills act as assets that appreciate in value, rather than just cost items.
Anthropic identified nine categories of skills, ranging from library references and data analysis to operational runbooks, with verification skills being the most impactful in improving output quality. The company advocates for focused development of these categories to fill organizational gaps.
Technical lessons include avoiding restating obvious instructions, emphasizing non-obvious, specific knowledge, and carefully designing trigger descriptions that match user requests precisely. Bundling real code and helper functions within skills is also highlighted as a key practice.
A Skill is a folder, not a prompt
Anthropic published what it learned running hundreds of Skills across its own engineering org. Read as a business memo, the point is bigger than a coding trick: this is how ad-hoc prompting becomes durable institutional capability — the SOPs your agents actually follow, versioned and shared.
“A Skill is just a clever markdown prompt you save in a file.”
A folder the agent can discover, read & run — instructions, scripts, references, templates, config & on-demand hooks.
The knowledge of how your organization actually operates can be captured, versioned, shared & executed — and the thing capturing it is a humble folder with a script and a gotchas list inside. For the builder, that’s context engineering with real tools attached. For whoever owns the budget, it’s the difference between AI that starts from zero every morning and an asset that compounds. Caveats: best practices are still evolving, checked-in Skills cost context, and curation beats accumulation. Start with one Skill, one gotcha, and the category that catches your mistakes.
Implications for AI Development and Business Operations
This approach shifts the focus from ad-hoc prompting to building structured, reusable assets that embed organizational knowledge directly into AI systems. It enhances consistency across outputs, accelerates training and onboarding, and creates a living library of capabilities that improves over time. For businesses, this means more reliable AI services and a way to institutionalize tribal knowledge, reducing reliance on individual expertise.
By framing Skills as folders with comprehensive content, Anthropic demonstrates a path toward more robust, maintainable AI systems that can adapt and improve, potentially setting a new standard in enterprise AI deployment.
AI development folder structure tools
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
From Prompt Engineering to Asset Building in AI Teams
Traditionally, AI teams have relied on prompt engineering—crafting specific instructions for each task—to guide model behavior. This method is often brittle, requiring repeated effort and lacking institutional memory. Anthropic’s new approach, based on running hundreds of Skills structured as folders, represents a significant departure, emphasizing reusable, version-controlled assets that encapsulate organizational procedures and knowledge.
This development builds on ongoing industry efforts to move beyond simple prompt tuning, aiming for more durable, scalable, and maintainable AI capabilities. The concept of Skills as folders aligns with broader trends in software engineering, such as modularity and asset versioning, applied to AI workflows.
“A Skill is not just a prompt saved in a text file; it’s a folder that contains instructions, scripts, and assets that the agent can discover and execute.”
— Anthropic engineer
Unclear Aspects of Scaling and Adoption
It is not yet clear how widely this approach will be adopted outside Anthropic or how easily other organizations can implement similar folder-based Skills. Details on tooling, integration challenges, and scalability across diverse enterprise environments remain to be seen. Additionally, the long-term impact on AI performance and maintenance costs is still being evaluated.
Next Steps for Broader Adoption and Tooling
Organizations are likely to experiment with building their own Skills libraries based on this model, focusing on filling organizational gaps identified through the nine-category framework. Further development of tooling to support Skills creation, versioning, and discovery will be critical. Anthropic may also publish more detailed guides or open-source components to facilitate wider adoption.
Research and industry discussions will explore how this model impacts AI reliability, compliance, and operational efficiency, shaping future best practices in enterprise AI deployment.
Key Questions
What exactly is a Skill in Anthropic’s framework?
A Skill is a folder containing instructions, scripts, reference documents, and configuration assets that an AI agent can discover, read, and execute to perform specific tasks reliably.
How does this approach improve AI consistency?
By bundling all relevant instructions and assets into a structured folder, Skills ensure that the AI performs tasks in a standardized way, reducing variability caused by ad-hoc prompting.
Can this method be applied outside Anthropic?
While promising, the approach’s scalability and ease of adoption depend on tooling, organizational complexity, and integration capabilities. Other organizations are expected to experiment with similar models.
What are the main categories of Skills identified?
They include library references, product verification, data analysis, business process automation, code scaffolding, code review, deployment, runbooks, and infrastructure operations.
What is the most impactful type of Skill according to Anthropic?
The verification Skills, which check the output quality and prevent mistakes, are considered the most valuable for improving reliability.
Source: ThorstenMeyerAI.com
