--- title: OWASP Flags Top AI Skill Risks in New Security Blueprint description: OWASP has released a finalized Top 10 list of security risks for agentic AI 'skills' — natural-language and coded add-ons for AI agents — alongside a new... canonical: https://daily.dev/posts/owasp-flags-top-ai-skill-risks-in-new-security-blueprint-w5ejv40de twitter:card: summary_large_image twitter:site: @dailydotdev og:type: website og:site_name: daily.dev og:title: OWASP Flags Top AI Skill Risks in New Security Blueprint | daily.dev og:description: OWASP has released a finalized Top 10 list of security risks for agentic AI 'skills' — natural-language and coded add-ons for AI agents — alongside a new... og:url: https://daily.dev/posts/owasp-flags-top-ai-skill-risks-in-new-security-blueprint-w5ejv40de og:image: https://api.daily.dev/og/posts/w5ejv40dE.png og:image:alt: OWASP Flags Top AI Skill Risks in New Security Blueprint og:image:width: 1200 og:image:height: 630 og:locale: en --- > ## Documentation Index > Fetch the complete documentation index at: https://daily.dev/llms.txt > Use this file to discover all available pages before exploring further. # OWASP Flags Top AI Skill Risks in New Security Blueprint **[Dark Reading](https://daily.dev/sources/dr)** · 7 min read · 0 upvotes · 0 comments ## Summary OWASP has released a finalized Top 10 list of security risks for agentic AI 'skills' — natural-language and coded add-ons for AI agents — alongside a new Universal Agentic Skill Format v1.0 YAML template to help standardize and secure skills. The list's top risks are Malicious Skills and Supply Chain Compromise, driven partly by a July incident where trojanized skills impersonating a platform called Paperclip racked up over 300,000 installs before detection. Experts from Microsoft and startup Air argue organizations need visibility into which agents and skills their workers use, since most CISOs can't currently account for their skill exposure or respond quickly to a malicious skill alert. ## Full article daily.dev links to this article rather than hosting it. Read it at the original source: ## Questions this post answers ### What is OWASP's Top 10 list for AI agent skills and what is the number one risk? OWASP released a finalized Agentic Skills Top 10 list covering two critical, four high-severity, and four medium-severity risks. Malicious Skills ranks number one, followed by Supply Chain Compromise at number two, reflecting how untrusted natural-language or coded add-ons for AI agents can be trojanized and distributed through repositories like GitHub. _Track emerging agentic AI security standards like this one on daily.dev as your team adopts skills._ ### What is the Universal Agentic Skill Format and what does it include? The Universal Agentic Skill Format v1.0 is a YAML template released by OWASP designed to standardize AI agent skills so automated tools can verify legitimacy and detect malicious ones. It includes sections for provenance, permissions for different actions, requirements and dependencies, signatures and hashes, and a changelog, aiming to separate user instructions from skill-supplied content. _Developers securing agent pipelines can follow format standards like this one on daily.dev._ ### How did attackers exploit a look-alike domain to spread malicious AI skills in the Paperclip incident? In early July, an attacker registered a look-alike domain impersonating the agentic AI platform Paperclip and distributed both trojanized Python packages and weaponized AI skills to steal credentials and sensitive information. While automated scanners caught the malicious Python packages within hours, the trojanized skills evaded detection and reached more than 300,000 installs each, according to research published August 6. _daily.dev helps security-minded developers stay ahead of supply chain attacks like this one._ ## Similar posts on daily.dev - [Managing agentic AI risk: Lessons from the OWASP Top 10](https://daily.dev/posts/managing-agentic-ai-risk-lessons-from-the-owasp-top-10-dkauj1rfw) · CSO Online · 1 upvotes · 0 comments - [OWASP Project Publishes List of Top Ten AI Agent Threats](https://daily.dev/posts/owasp-project-publishes-list-of-top-ten-ai-agent-threats-8gtojufml) · Security Boulevard · 0 upvotes · 0 comments - [All About OWASP Top 10 for Agentic AI Applications](https://daily.dev/posts/all-about-owasp-top-10-for-agentic-ai-applications-kmg1xmz5w) · InfoSec Write-ups · 2 upvotes · 0 comments --- Tags: [#security](https://daily.dev/tags/security), [#mcp](https://daily.dev/tags/mcp), [#ai-security](https://daily.dev/tags/ai-security), [#agentic-ai](https://daily.dev/tags/agentic-ai) [View this post on daily.dev](https://daily.dev/posts/owasp-flags-top-ai-skill-risks-in-new-security-blueprint-w5ejv40de) ```json {"@context":"https://schema.org","@graph":[{"@type":"Organization","@id":"https://daily.dev/#organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180},"sameAs":["https://twitter.com/dailydotdev","https://github.com/dailydotdev","https://www.linkedin.com/company/daily-dev-ltd"]},{"@type":"WebSite","@id":"https://daily.dev/#website","url":"https://daily.dev","name":"daily.dev","publisher":{"@id":"https://daily.dev/#organization"},"potentialAction":{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https://daily.dev/search?q={search_term_string}"},"query-input":"required name=search_term_string"}}]} {"@context":"https://schema.org","@type":"TechArticle","headline":"OWASP Flags Top AI Skill Risks in New Security Blueprint","url":"https://daily.dev/posts/owasp-flags-top-ai-skill-risks-in-new-security-blueprint-w5ejv40de","mainEntityOfPage":{"@type":"WebPage","@id":"https://daily.dev/posts/owasp-flags-top-ai-skill-risks-in-new-security-blueprint-w5ejv40de"},"datePublished":"2026-08-21T17:46:11.450Z","dateModified":"2026-08-23T04:54:57.906Z","description":"OWASP has released a finalized Top 10 list of security risks for agentic AI 'skills' — natural-language and coded add-ons for AI agents — alongside a new...","image":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/165f9309f29ada5d4d43c223cd80baa1?_a=AQAEuop","thumbnailUrl":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/165f9309f29ada5d4d43c223cd80baa1?_a=AQAEuop","isAccessibleForFree":true,"articleSection":"Dark Reading","inLanguage":"en","publisher":{"@type":"Organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180}},"author":{"@type":"Organization","name":"Dark Reading","logo":"https://media.daily.dev/image/upload/t_logo,f_auto/v1/logos/dr","url":"https://daily.dev/sources/dr"},"commentCount":0,"discussionUrl":"https://daily.dev/posts/owasp-flags-top-ai-skill-risks-in-new-security-blueprint-w5ejv40de","interactionStatistic":[{"@type":"InteractionCounter","interactionType":{"@type":"LikeAction"},"userInteractionCount":0},{"@type":"InteractionCounter","interactionType":{"@type":"CommentAction"},"userInteractionCount":0}],"keywords":"security,mcp,ai-security,agentic-ai","timeRequired":"PT7M"} {"@context":"https://schema.org","@type":"BreadcrumbList","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https://daily.dev"},{"@type":"ListItem","position":2,"name":"Dark Reading","item":"https://daily.dev/sources/dr"},{"@type":"ListItem","position":3,"name":"OWASP Flags Top AI Skill Risks in New Security Blueprint"}]} {"@context":"https://schema.org","@type":"FAQPage","@id":"https://daily.dev/posts/owasp-flags-top-ai-skill-risks-in-new-security-blueprint-w5ejv40de#faq","mainEntity":[{"@type":"Question","name":"What is OWASP's Top 10 list for AI agent skills and what is the number one risk?","acceptedAnswer":{"@type":"Answer","text":"OWASP released a finalized Agentic Skills Top 10 list covering two critical, four high-severity, and four medium-severity risks. Malicious Skills ranks number one, followed by Supply Chain Compromise at number two, reflecting how untrusted natural-language or coded add-ons for AI agents can be trojanized and distributed through repositories like GitHub. Track emerging agentic AI security standards like this one on daily.dev as your team adopts skills."}},{"@type":"Question","name":"What is the Universal Agentic Skill Format and what does it include?","acceptedAnswer":{"@type":"Answer","text":"The Universal Agentic Skill Format v1.0 is a YAML template released by OWASP designed to standardize AI agent skills so automated tools can verify legitimacy and detect malicious ones. It includes sections for provenance, permissions for different actions, requirements and dependencies, signatures and hashes, and a changelog, aiming to separate user instructions from skill-supplied content. Developers securing agent pipelines can follow format standards like this one on daily.dev."}},{"@type":"Question","name":"How did attackers exploit a look-alike domain to spread malicious AI skills in the Paperclip incident?","acceptedAnswer":{"@type":"Answer","text":"In early July, an attacker registered a look-alike domain impersonating the agentic AI platform Paperclip and distributed both trojanized Python packages and weaponized AI skills to steal credentials and sensitive information. While automated scanners caught the malicious Python packages within hours, the trojanized skills evaded detection and reached more than 300,000 installs each, according to research published August 6. daily.dev helps security-minded developers stay ahead of supply chain attacks like this one."}}]} ```