hyper-browsing
Fail
Audited by Snyk on Nov 19, 2026
Risk Level: CRITICAL
Full Analysis
CRITICAL E004: Prompt injection detected in skill instructions.
- Potential prompt injection detected (medium risk: 0.30). The skill is a documented access-fallback workflow, but the reader recipe automatically sends the user's
JINA_API_KEYas an Authorization header to a third-party reader, which contradicts the skill's own forbidden rule against sending Authorization to readers without an explicit request.
MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).
- Third-party content exposure detected (low risk: 0.10). The workflow reads outsider-authored text (arbitrary web pages, public-platform posts, reader and archive copies) only after the agent fetches a specific URL, handle, or endpoint the user named or a search returned, so the attacker can publish poison but cannot force delivery.
Issues (2)
E004
CRITICALPrompt injection detected in skill instructions.
W011
MEDIUMThird-party content exposure detected (indirect prompt injection risk).
Audit Metadata