atak-przeciwnika-pl
Pass
Audited by Gen Agent Trust Hub on Oct 9, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and analyze user-provided legal documents such as pleadings, appeals, and opinions, creating a surface for indirect prompt injection where instructions embedded in the documents could attempt to influence the agent's behavior.
- Ingestion points: The 'Co dostarcza użytkownik' section specifies that the skill processes external legal texts, appeals, and reasoning fragments.
- Boundary markers: The skill lacks explicit delimiters for user-provided text in its internal instructions, though it recommends the use of a companion skill ('let-it-be') for pseudonymization before analysis.
- Capability inventory: The skill's execution environment is restricted via YAML frontmatter to the 'Read' tool only, preventing unauthorized network operations or file modifications.
- Sanitization: The instructions do not define explicit sanitization, filtering, or escaping protocols for the ingested legal content.
Audit Metadata