ochrana-danych-uczniow
Pass
Audited by Gen Agent Trust Hub on Nov 14, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill exhibits an attack surface for indirect prompt injection because it is designed to process and audit untrusted external data such as school application source code, database exports, and student records.\n
- Ingestion points: Untrusted data enters the context through school application code being audited, student data records (CSV, SQL), and pedagogical documentation (found in SKILL.md, references/data.md, and references/storage.md).\n
- Boundary markers: Present. Instructions explicitly mandate the use of synthetic data for testing and prohibit requesting production passwords or real student data until formal authorization (found in SKILL.md).\n
- Capability inventory: The skill allows the agent to perform security audits, implement cryptographic logic, modify authentication flows (SSO/RBAC), and handle file exports (found in SKILL.md, references/identity.md, and references/crypto.md).\n
- Sanitization: Present. The skill provides specific instructions to sanitize CSV exports to prevent formula injection and requires the use of parameterized queries to prevent SQL injection (found in references/storage.md).\n- [SAFE]: The skill references various external legal and technical resources from official government domains (sejm.gov.pl, uodo.gov.pl, boe.es) and well-known organizations (EDPB, NIST, GitHub). These references are documented neutrally and align with the skill's primary purpose of providing data protection guidance.
Audit Metadata