ads-landing
Pass
Audited by Gen Agent Trust Hub on Oct 17, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted content from external landing pages, creating a potential surface for indirect prompt injection where instructions could be hidden in audited pages.
- Ingestion points: The skill is designed to ingest and audit the content of landing pages, redirects, frames, and scripts (identified in
SKILL.md). - Boundary markers: The instructions explicitly mandate treating all external content as untrusted and using a "guarded HTTP fetcher" with egress-sandbox attestations.
- Capability inventory: The skill involves network fetching operations to retrieve external content.
- Sanitization: A detailed "blocked-navigation contract" is defined to validate URLs and prevent access to private, loopback, link-local, and cloud-metadata IP addresses, effectively mitigating SSRF risks.
Audit Metadata