📊 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 shifted from simple prompts to packaging knowledge into reusable ‘Skills’ structured as folders. This approach improves consistency, onboarding, and asset development in AI workflows, marking a significant departure from traditional prompt engineering.
Anthropic has announced that its AI Skills are best understood as folders containing instructions, scripts, and assets, rather than simple prompts. This shift aims to improve consistency, onboarding, and institutional memory within AI teams, marking a significant evolution in how organizations deploy AI capabilities. The insight comes from a detailed write-up by a Claude Code engineer, reflecting Anthropic’s internal practices and lessons learned from running hundreds of Skills across its engineering organization.
According to Anthropic, a Skill is a folder that can hold instructions, reference documents, scripts, templates, data, configuration, and hooks. This structure allows AI agents to discover and execute complex workflows as a cohesive unit, rather than relying on ad-hoc prompts. The approach shifts the focus from prompts as fleeting instructions to Skills as durable, versioned assets that encode organizational knowledge.
Anthropic emphasizes that Skills improve operational consistency — ensuring the same task produces similar results regardless of who runs it — and facilitate onboarding by capturing tribal knowledge in a reusable format. The company also notes that Skills can evolve and compound over time, becoming more refined through continuous use and iteration. The internal taxonomy identifies nine categories of Skills, from library references to infrastructure operations, with verification Skills deemed most valuable for quality control.
Technical lessons highlight that effective Skills avoid restating obvious information, instead focusing on non-obvious, specific guidance. The description of each Skill functions as a trigger for the agent, requiring precise wording to activate the correct workflow. Bundling real code and helper functions within Skills enhances their utility, making them more than just static instructions.
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 Organizational AI Deployment
This development signals a shift toward more structured, maintainable, and scalable AI workflows within organizations. By treating Skills as containers of institutional knowledge, companies can achieve greater consistency, reduce onboarding time, and develop a library of reusable, improving assets. This approach also enables teams to move beyond simple prompt engineering toward building durable operational capabilities, potentially transforming AI’s role in enterprise workflows.
AI for Project Managers: A Desk Reference & Field Guide: Use Artificial Intelligence to Streamline Workflows, Automate Tasks, and Make Smarter Decisions with Practical Tools and Ethical Insights
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
From Prompting to Asset Management in AI Development
Traditional AI deployment often relies on ad-hoc prompting, requiring teams to craft and re-craft instructions for each task. Anthropic’s internal experience shows that packaging knowledge into Skills as folders represents a fundamental shift, inspired by principles of asset management and institutional memory. The concept aligns with broader trends in AI engineering, emphasizing modularity, version control, and reusable components, similar to software engineering practices. The approach builds on prior efforts to improve AI reliability but elevates it by structuring workflows as durable assets rather than ephemeral prompts.
Anthropic’s insights come from running hundreds of Skills internally, revealing patterns and categories that serve as a blueprint for other organizations seeking to institutionalize AI capabilities effectively.
“A Skill is a folder — one that can contain instructions, reference documents, scripts, templates, data, configuration, and even hooks that fire only while the Skill is active.”
— Thorsten Meyer, AI Engineer at Anthropic
![MixPad Free Multitrack Recording Studio and Music Mixing Software [Download]](https://m.media-amazon.com/images/I/71ltIxIuz1L._SL500_.jpg)
MixPad Free Multitrack Recording Studio and Music Mixing Software [Download]
- Multitrack Recording and Mixing: Create mixes with audio, music, and voice tracks
- Track Customization: Apply effects and editing tools to tracks
- Music Creation Tools: Includes Beat Maker and MIDI Creator
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Unresolved Questions About Skill Adoption and Scaling
It is not yet clear how widely organizations will adopt this folder-based approach outside Anthropic or how it scales in different enterprise contexts. The specific technical challenges of maintaining and updating large Skills libraries, as well as integration with existing systems, remain to be seen. Additionally, the impact on AI transparency and explainability, given the complexity of folder contents, is still under discussion.
AI and UX: Why Artificial Intelligence Needs User Experience
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Future Steps for Implementing Folder-Based Skills
Organizations interested in this approach should evaluate their current workflows and identify categories of Skills that could improve consistency and onboarding. Further research and pilot projects are expected to explore how to best structure, version, and maintain Skills libraries at scale. Anthropic may also release tools or best practices to facilitate broader adoption of folder-based Skills in enterprise AI deployments.
Plaud Note AI Voice Recorder, Note Taker w/Case, App Control, Transcribe & Summarize with AI, Support 112 Languages, for Meetings, Calls, Lectures, Professionals, Teams, Black, Non-Pro Version
- Language Support: Captures 112 languages with AI transcription
- AI-Powered Insights: Generates transcripts, mind maps, to-do lists
- Slim and Lightweight: Only 0.12 inches thick, 1.06 oz
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Key Questions
How does treating Skills as folders differ from traditional prompt engineering?
Unlike prompts, which are single instructions or questions, Skills as folders contain multiple assets, including instructions, scripts, and data, making them more durable, reusable, and capable of encapsulating complex workflows.
What benefits does this approach offer for organizations?
It improves output consistency, reduces onboarding time by capturing tribal knowledge, and creates a library of assets that can be refined and scaled over time, enhancing organizational AI capabilities.
Are there technical challenges associated with implementing Skills as folders?
Yes, maintaining and updating large Skills libraries, ensuring proper trigger descriptions, and integrating with existing systems can pose challenges, especially at scale. These are areas for ongoing development and best practice formulation.
Will this approach work with all types of AI tasks?
The approach is most suited for operational and repetitive tasks where consistency and institutional knowledge are critical. Its applicability to more creative or less structured tasks remains to be explored.
Source: ThorstenMeyerAI.com
