A Skill Is a Folder, Not a Prompt: What Anthropic Learned Running Hundreds of Them

📊 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 the understanding of AI Skills from prompts to folders containing instructions, scripts, and references. This approach enhances consistency, onboarding, and institutional knowledge. The company ran hundreds of these Skills internally, demonstrating their value as organizational assets.

Anthropic has redefined what constitutes an AI Skill, emphasizing that it is a folder rather than a prompt. This new understanding stems from the company’s experience running hundreds of Skills internally, which has implications for how organizations develop, share, and maintain AI capabilities. The approach aims to create durable, reusable assets that improve consistency, onboarding, and institutional memory within AI teams.

According to a write-up from an Anthropic Claude Code engineer, a Skill is not simply a saved prompt but a folder containing instructions, reference documents, scripts, templates, data, and configuration. The agent can discover and execute the contents of this folder, making Skills more like containers of organizational knowledge than static prompts.

Anthropic’s internal experience shows that Skills help standardize output, reduce onboarding time, and improve over time as they are refined through repeated use. The company identified nine categories of Skills, ranging from library references and product verification to infrastructure operations, with verification being the most impactful.

Building effective Skills involves avoiding redundancy, focusing on non-obvious, specific instructions, and including ‘Gotchas’—traps or pitfalls that the agent must avoid. The description of each Skill acts as a trigger, matching user requests with the appropriate Skill based on language and internal slang, ensuring the agent activates the right container for the task.

At a glance
reportWhen: published recently, based on internal A…
The developmentAnthropic published insights from its internal experience running hundreds of Skills, emphasizing that Skills are folders, not prompts, to improve AI agent reliability and organizational knowledge sharing.
Crypto market snapshot
Fear & Greed Index
24/100 — Extreme Fear
Bitcoin BTC$63,040▲ 0.6%
Ethereum ETH$1,772▲ 0.5%
Tether USDT$0.999▼ 0.0%
BNB BNB$582.54▲ 2.1%
USDC USDC$0.9998▼ 0.0%
XRP XRP$1.14▲ 0.4%
Solana SOL$80.38▲ 0.0%
TRON TRX$0.3291▲ 1.4%
Live data · CoinGecko · alternative.me (24h change)
A Skill Is a Folder, Not a Prompt — Insights
AI Dispatch · Insights · 1 July 2026

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.

✕ The misconception

“A Skill is just a clever markdown prompt you save in a file.”

✓ What it actually is

A folder the agent can discover, read & run — instructions, scripts, references, templates, config & on-demand hooks.

Anatomy of a Skill — the file system is context engineering
my-skill/the unit you share & version
├─ SKILL.mdroot instructions + a description written for the model (its trigger)
├─ references/deep detail pulled in only when needed — progressive disclosure
├─ scripts/real code, so the agent composes instead of rebuilding boilerplate
├─ assets/templates & files to copy into the output
├─ config.jsonsetup the agent asks for if it’s missing (e.g. which Slack channel)
└─ hooks + memoryon-demand guardrails + an append-only log so it remembers
Why it matters: the folder itself is the knowledge base. The agent reads the root, then reaches deeper only when the task demands it — the same way you’d hand a new hire a one-pager that points to the detailed docs.
The nine types — a gap-analysis map for your own library
1Library / API reference
2Product verification ★ top impact
3Data fetching & analysis
4Business-process automation
5Code scaffolding & templates
6Code quality & review
7CI/CD & deployment
8Runbooks
9Infrastructure operations
By Anthropic’s own measurement, verification Skills — the ones that check the work — moved output quality the most. If you build one category well, build that one.
The craft — what separates a good Skill from a useless one
Gotchas = highest-signal section Describe for the model, not humans (it’s the trigger) Don’t state the obvious Ship scripts, not just prose On-demand guardrail hooks (/careful, /freeze) Let it remember (log / SQLite) Don’t railroad — leave room to adapt
The take

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.

Source: “Lessons from building Claude Code: How we use skills,” Thariq Shihipar (Anthropic), Claude blog, 3 June 2026. Categories, examples & measured claims are Anthropic’s; framing is the author’s. Docs: code.claude.com/docs/en/skills.
thorstenmeyerai.com

Implications for AI Development and Organizational Knowledge

This approach transforms how companies build and maintain AI capabilities, turning Skills into valuable, evolving assets that encode organizational procedures, tribal knowledge, and best practices. It shifts the focus from ad-hoc prompt engineering to durable, reusable containers, potentially improving consistency, reducing onboarding costs, and enabling continuous refinement. For organizations investing heavily in AI automation, this method could lead to more reliable and scalable systems, but it also requires a cultural shift toward managing Skills as organizational assets.
Amazon

AI development folder templates

As an affiliate, we earn on qualifying purchases.

As an affiliate, we earn on qualifying purchases.

From Prompt Engineering to Containerized Skills in AI Workflows

Until now, most teams using AI coding agents relied on repeated prompt engineering—crafting and reusing prompts for different tasks. Anthropic’s internal experience suggests that this approach is limited in scalability and durability. The company’s recent documentation emphasizes that Skills, as folders containing instructions and scripts, are a more effective way to embed organizational knowledge into AI systems.

Anthropic’s insights come after running hundreds of Skills internally, allowing the company to identify nine core categories and refine their design principles. This move aligns with broader trends toward institutionalizing AI capabilities and making them part of standard operating procedures rather than ad-hoc prompts.

“A Skill is a folder, not just a prompt. It’s a container for instructions, scripts, and knowledge that the agent can discover and execute.”

— Thorsten Meyer, AI researcher at Anthropic

Amazon

AI instruction and script management tools

As an affiliate, we earn on qualifying purchases.

As an affiliate, we earn on qualifying purchases.

Unclear Aspects of Skill Implementation and Scaling

While Anthropic’s internal experience demonstrates the benefits of folder-based Skills, it is not yet clear how easily this approach can be adopted by other organizations or scaled across different AI systems. Details about the specific processes for creating, managing, and updating Skills at scale remain limited, and the long-term effectiveness of this method is still being evaluated.

Amazon

AI knowledge base organization software

As an affiliate, we earn on qualifying purchases.

As an affiliate, we earn on qualifying purchases.

Next Steps for Broader Adoption and Refinement

Organizations interested in this approach should begin cataloging their internal procedures into Skills, focusing on non-obvious instructions and ‘Gotchas.’ Future developments may include tooling for easier creation, versioning, and sharing of Skills across teams. Anthropic is likely to continue refining its internal process and share lessons learned to facilitate wider adoption.

Amazon

AI agent configuration folders

As an affiliate, we earn on qualifying purchases.

As an affiliate, we earn on qualifying purchases.

Key Questions

How is a Skill different from a prompt?

A Skill is a folder containing instructions, scripts, and knowledge, acting as a reusable container for organizational procedures, rather than a one-time prompt or instruction text.

What benefits does this approach offer?

Skills improve output consistency, reduce onboarding time, and enable continuous refinement, making AI systems more reliable and aligned with organizational processes.

Can this method be applied outside of AI coding agents?

Yes, the concept of containerized, reusable organizational assets can potentially be adapted to other AI applications and operational workflows.

What challenges might organizations face in adopting this approach?

Challenges include developing effective Skills, managing version control, and integrating this method into existing workflows without excessive overhead.

Will Skills replace prompt engineering entirely?

Skills aim to supplement prompt engineering by creating more durable, maintainable assets, but prompt design may still play a role in specific contexts.

Source: ThorstenMeyerAI.com

Nothing in this article is financial or investment advice. Cryptocurrency and precious-metal investments carry significant risk — do your own research and consider a licensed advisor.
You May Also Like

RHEO: Paint With Light

RHEO is a simple, beautifully designed app that transforms touch into flowing light art on iPhone, iPad, and Apple Vision Pro, emphasizing calm and accessibility.

The Compute Reckoning: Anthropic Finally Admits What Customers Suspected for Ten Months

Anthropic reveals that its recent customer experience issues were due to compute shortages, now addressed through a major partnership with SpaceX and other capacity expansions.

The bridge. Why the AI buildout runs on a nuclear story and a gas reality.

Exploring the gap between AI data center power needs, nuclear procurement, and the current reliance on natural gas infrastructure.

The OAuth Permission Apocalypse.

An analysis of the growing risk from overly permissive OAuth permissions, exemplified by recent supply chain breaches, and why industry-wide change is urgent.