web-development

Pass

Audited by Gen Agent Trust Hub on Oct 23, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill instructs the agent to use browser automation tools to verify web applications. This creates an attack surface where the agent processes untrusted content from the browser while possessing powerful management capabilities.
  • Ingestion points: The agent-browser tool is used to load and interact with local or remote web pages (detailed in browser-testing.md).
  • Boundary markers: No specific instructions are provided to the agent to disregard instructions potentially embedded in the HTML or DOM of the pages it browses.
  • Capability inventory: The skill utilizes several MCP tools for infrastructure management, including manageHosting, queryHosting, manageApps, manageCloudRun, and manageFunctions (referenced in SKILL.md and frameworks.md).
  • Sanitization: No sanitization or filtering is mentioned for content ingested via the browser tool.
  • [EXTERNAL_DOWNLOADS]: The skill provides a link to an external JavaScript SDK hosted on a Content Delivery Network.
  • Evidence: https://static.cloudbase.net/cloudbase-js-sdk/latest/cloudbase.full.js referenced in SKILL.md.
  • Context: The resource is hosted on the official vendor domain for CloudBase integration.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 23, 2026, 06:19 AM