hierarchia-zrodel-pl
Pass
Audited by Gen Agent Trust Hub on Oct 12, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to process untrusted user-supplied legal queries to generate research maps, which presents a surface for instructions embedded in data to influence the agent's behavior.
- Ingestion points: Untrusted data enters the agent context through legal questions or descriptions of legal situations provided by the user (as described in the 'Kiedy używać' section of
SKILL.md). - Boundary markers: The skill uses
[DO SPRAWDZENIA]as a marker for unverified information in its output template, though it lacks explicit delimiters to wrap user-provided context to prevent instruction leakage. - Capability inventory: The skill is configured with the
Readtool in theallowed-toolsmetadata, limiting its direct action capabilities. - Sanitization: No explicit sanitization, escaping, or validation of user-provided legal scenarios is implemented within the prompt instructions.
Audit Metadata