israeli-insurance-comparator
Pass
Audited by Gen Agent Trust Hub on May 15, 2026
Risk Level: SAFENO_CODE
Full Analysis
- [SAFE]: The skill's functionality is consistent with its stated purpose. Analysis of instructions and metadata revealed no malicious code, obfuscation, or unauthorized access patterns.
- [NO_CODE]: No scripts, binaries, or automated command sequences are provided. The skill relies entirely on natural language instructions for the agent to use available tools manually.
- [DATA_EXFILTRATION]: The skill directs the agent to collect sensitive Personally Identifiable Information (PII), specifically Israeli ID numbers, which are required for legitimate insurance quoting processes. The skill includes guidance for users to only provide this information on secure, regulated platforms.
- [PROMPT_INJECTION]: The skill presents a surface for indirect prompt injection through external data ingestion. 1. Ingestion points: Insurance comparison websites (e.g., car.cma.gov.il, wobi.co.il) accessed via WebFetch. 2. Boundary markers: None specified in instructions. 3. Capability inventory: Bash and WebFetch tools are available. 4. Sanitization: No explicit sanitization or validation of fetched content is performed. This surface is considered safe given the well-known nature of the referenced services.
Audit Metadata