first-party-data-collection
Pass
Audited by Gen Agent Trust Hub on Sep 16, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill describes and tests the implementation of systems that ingest untrusted data from external users (quizzes, surveys, and preference centers). This creates a vulnerability surface where malicious input could influence agent behavior if processed without sanitization.\n
- Ingestion points: Customer profile data collected via quizzes and forms as described in
SKILL.mdandevals/customer-schema-and-progressive-profilin/task.md.\n - Boundary markers: The provided tasks and instructions do not explicitly implement or require boundary markers to delimit untrusted data.\n
- Capability inventory: The system is designed to perform database updates (
db.customerProfiles.update), notify email service providers (emailServiceClient.updateContactProperty), and interact with ad platforms.\n - Sanitization: There is no explicit mention of sanitization or validation of the ingested user data in the provided instructions or evaluation criteria.
Audit Metadata