40Medium
The skill facilitates SSRF via arbitrary URL fetching and enables remote command injection by piping unverified API responses directly into shell commands, violating its stated network access restrictions.
Skill Info
Namesleekdotdesign/agent-skills/sleek-design-mobile-apps
Registrygithub
Version86079671c159
PURLpkg:github/sleekdotdesign/agent-skills@86079671c159?skill=sleek-design-mobile-apps
Stars426
Installs249,300
Source
Install
This skill has findings worth reviewing before installing.
GitHub
248,900 installs
Skills.shDocs
npx skills add https://github.com/sleekdotdesign/agent-skillsAssessments (3)
AI Agent Traps ↗Description Mismatchmediumdescription mismatchAML.T0011.000LLM06:2025LLM09:2025
The skill documentation claims network access is limited to 'https://sleek.design only', but the implementation instructions explicitly require fetching icons from 'https://api.iconify.design'.
85% confidence
The 'Prerequisites' section states: 'Network access limited to https://sleek.design only.' However, the 'Icons' section instructs: 'fetch the SVGs from the Iconify API... GET https://api.iconify.design/{prefix}/{name}.svg'.Potential for SSRF via imageUrlsmediumresource abuseAML.T0029AML.T0034LLM10:2025
The skill allows passing arbitrary HTTPS URLs to the Sleek API, which are then fetched by Sleek's servers. This creates a potential Server-Side Request Forgery (SSRF) vector if the user provides internal or sensitive URLs.
60% confidenceline 189
"imageUrls": ["https://example.com/ref.png"]
Potential for indirect prompt injection via icon fetchingmediumdata exfiltrationAML.T0025AML.T0024LLM02:2025
The skill instructs the agent to fetch SVGs from an external, unverified API (api.iconify.design) based on icon names found in the HTML code. An attacker could manipulate the component HTML to include malicious URLs or trigger requests to arbitrary domains.
70% confidenceline 475
GET https://api.iconify.design/{prefix}/{name}.svgEncouragement of direct shell-based file writingmediumcommand executionAML.T0050AML.T0053AML.T0072LLM05:2025LLM06:2025
The skill explicitly instructs the agent to use shell commands to pipe API responses directly into files, bypassing standard agent-controlled file writing tools and increasing the risk of arbitrary command injection if the API response is manipulated.
80% confidenceline 505
use shell commands to fetch the API response and write it directly to disk
Missing licenseinfopolicy violation
Skill does not specify a license field. Specifying a license helps users understand usage terms.
100% confidence
Badge
Markdown
[](https://mondoo.com/ai-agent-security/skills/github/sleekdotdesign/agent-skills/sleek-design-mobile-apps)HTML
<a href="https://mondoo.com/ai-agent-security/skills/github/sleekdotdesign/agent-skills/sleek-design-mobile-apps"><img src="https://mondoo.com/ai-agent-security/api/badge/github/sleekdotdesign/agent-skills/sleek-design-mobile-apps.svg" alt="Mondoo Skill Check" /></a>Image URL
https://mondoo.com/ai-agent-security/api/badge/github/sleekdotdesign/agent-skills/sleek-design-mobile-apps.svgSecure your AI agents
Skills can read files, run commands, and access credentials. Mondoo helps organizations manage the security risks of AI agent skills across their entire fleet.
- Continuous scanning of skills across all registries
- Policy enforcement before skills reach your agents
- Integration with your existing security workflow