browser-extension-builder
Pass
Audited by Gen Agent Trust Hub on Nov 1, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill consists of instructional content and code templates for browser extension development. It adheres to standard practices and includes a section on anti-patterns to discourage insecure development habits such as requesting unnecessary permissions.
- [INDIRECT_PROMPT_INJECTION]: The skill provides templates for extensions that process untrusted data from web pages, which is an inherent attack surface for browser extensions. * Ingestion points: Example code in SKILL.md (content scripts) shows how to read text from web pages using document.querySelector. * Boundary markers: The provided code snippets do not include delimiters or instructions to treat web-injected data as untrusted. * Capability inventory: The templates demonstrate the use of chrome.storage for data persistence and DOM manipulation for UI injection. * Sanitization: The examples do not include logic for sanitizing or validating data retrieved from the web page before processing or storage.
Audit Metadata