israeli-building-committee

Pass

Audited by Gen Agent Trust Hub on Oct 14, 2026

Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTIONNO_CODE
Full Analysis
  • [COMMAND_EXECUTION]: The skill includes an optional Python script scripts/fee_split.py intended for hosts that support local execution. Analysis of the script confirms it is a standard-library-only utility that performs basic arithmetic to calculate building fee ratios. It contains no network requests, file system writes (outside of potential log output), or obfuscated commands.
  • [INDIRECT_PROMPT_INJECTION]: The skill is designed to process user-supplied data such as apartment areas and building bylaws (takanon). While this creates an ingestion surface for untrusted data, the skill includes explicit legal disclaimers and instructions for the agent to ask for specific clauses rather than making legal determinations. There are no high-privilege capabilities (like arbitrary code execution or network exfiltration) to exploit via this vector, and the primary output consists of factual summaries and Hebrew document templates.
  • [EXTERNAL_DOWNLOADS]: The skill references several authoritative sources, including Wikisource (he.wikisource.org), Kol Zchut (kolzchut.org.il), and official government services (gov.il). These are well-known, trusted organizations and information providers in the Israeli legal domain. References to these sites are informative and do not trigger a verdict escalation.
  • [NO_CODE]: Outside of the helper script, the skill operates entirely through natural language instructions and worked mathematical examples, minimizing the technical attack surface.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 14, 2026, 08:27 AM