prototype

Pass

Audited by Gen Agent Trust Hub on Jul 11, 2026

Risk Level: SAFE
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill ingests data from local project files (DESIGN.md and JOURNEY.md) and incorporates the content into a generated HTML file which is then processed by a subagent via screenshot. This creates a standard indirect injection surface where content in the input files could theoretically attempt to influence the subagent's summary, though no malicious behavior is directed by the skill itself.
  • [DATA_EXPOSURE]: The skill uses the browser MCP to render local files via the file:// protocol. While this involves enabling the allow_internal permission, the skill is structured to navigate only to the specific output paths it generates for the prototype artifacts, which is a required function for the skill's visualization purpose.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 11, 2026, 03:18 AM
Security Audit — agent-trust-hub — prototype