sleek-design-mobile-apps
Pass
Audited by Gen Agent Trust Hub on Sep 17, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill represents a surface for indirect prompt injection through the following evidence chain:
- Ingestion points: Untrusted data enters the agent context via the
GET /api/v1/projects/:id/components/:componentIdendpoint, which returns HTML code for design components. - Boundary markers: There are no explicit instructions to the agent to ignore or delimit embedded instructions within the fetched HTML content.
- Capability inventory: The skill has the capability to execute shell commands (e.g., using
curlto fetch and pipe data) and write content to the local filesystem (saving.htmlfiles). - Sanitization: No sanitization or validation of the fetched HTML content is performed before it is written to the disk or processed.
- Ingestion points: Untrusted data enters the agent context via the
- [COMMAND_EXECUTION]: The skill recommends using shell commands to handle large API responses. Specifically, it suggests piping
curloutput directly into files to optimize performance and token usage when saving component HTML. - [EXTERNAL_DOWNLOADS]: The skill fetches resources from external services:
- Downloads visual assets from the
sleekdotdesignGitHub repository. - Fetches SVG icons from the Iconify API (
api.iconify.design), which is a well-known service for design icons.
- Downloads visual assets from the
Audit Metadata