signup-flow-cro
Pass
Audited by Gen Agent Trust Hub on Sep 25, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests external data from web searches and scraped pages, creating a vulnerability surface for indirect prompt injection.
- Ingestion points: The skill reads from approved public or connected data sources, including scraped page text and reviews as defined in
SKILL.mdandreferences/connectors.md. - Boundary markers: The skill includes explicit instructions in
SKILL.mdandreferences/skill-contract.mdto treat external data as untrusted input and specifically warns not to follow embedded instructions. - Capability inventory: Instructions limit the skill to generating recommendations by default, requiring explicit user authorization for any persistent file or account mutations.
- Sanitization: The
Unified Skill Contractrequires labeling all evidence with its source and confidence level to prevent misinterpretation of external data.
Audit Metadata