lgpd-consent-schema
Pass
Audited by Gen Agent Trust Hub on Nov 11, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: No security issues detected. The skill provides legitimate technical assets for developers to implement privacy-compliant data architectures.
- [DATA_EXPOSURE]: The skill handles potentially sensitive data such as IP addresses and User Agents (as evidence of consent) and includes a schema for storing encrypted CPFs (Brazilian Tax IDs). However, these are implemented following security best practices, such as application-level encryption and blind indexing, to minimize exposure.
- [COMMAND_EXECUTION]: The skill includes a PostgreSQL script to create triggers that enforce append-only behavior on the database level. This is a security-enhancing feature designed to prevent the modification or deletion of audit logs, which is a requirement for the burden of proof under LGPD.
- [INDIRECT_PROMPT_INJECTION]: While the TypeScript helpers process external data from request headers (e.g.,
x-forwarded-for,user-agent), the logic is strictly focused on database operations and hashing for data integrity. No capabilities for subprocess execution or external network operations were found that could be abused via malicious header content.
Audit Metadata