brand-content-strategy

Warn

Audited by Snyk on Jul 10, 2026

Risk Level: MEDIUM
Full Analysis

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

  • Third-party content exposure detected (high risk: 0.75). Yes: the required workflow step “Profile the brand” fetches an arbitrary website at runtime (WebFetch) and extracts verbatim text (positioning/credentials/CTAs) into the agent’s LLM context, which is outsider-authored free text from the target domain.

Issues (1)

W011
MEDIUM

Third-party content exposure detected (indirect prompt injection risk).

Audit Metadata
Risk Level
MEDIUM
Analyzed
Jul 10, 2026, 01:40 AM
Issues
1