lgpd-compliance
Pass
Audited by Gen Agent Trust Hub on Oct 31, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill utilizes shell commands with
rg(ripgrep) to search for specific PII and sensitive data patterns within codebases. This functionality is essential for its primary purpose of auditing software for compliance. - [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted user-provided content such as repositories and documents.
- Ingestion points: Code repositories, diffs, and privacy-related documents analyzed during the scoping and inventory phases.
- Boundary markers: Instructions explicitly require the agent to treat all artifacts strictly as evidence and to exclude raw data values from the final report.
- Capability inventory: The skill executes file-system searches and writes reports to the local environment.
- Sanitization: The agent is directed to describe data by category and location rather than including actual sensitive content, which mitigates the risk of processing malicious instructions embedded in the data.
- [SAFE]: URL references target official Brazilian government domains for legal and regulatory information, which are recognized as reliable sources.
Audit Metadata