swedish-industry-restaurang

Pass

Audited by Gen Agent Trust Hub on Oct 15, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The analysis identifies no security risks. The skill consists entirely of markdown documentation providing professional guidance on Swedish accounting standards (BAS 2026) and legal compliance (e.g., Alkohollagen, Skatteförfarandelag). No executable scripts, network requests, or unauthorized file access patterns were detected.- [PROMPT_INJECTION]: The instruction 'Always use over training data' is used to prioritize the specific regulatory information provided for the Swedish market over general AI knowledge. This is a common instructional pattern for RAG-based skills and does not attempt to bypass safety filters or override agent behavior beyond the intended domain focus.- [INDIRECT_PROMPT_INJECTION]: The skill mentions the processing of external data sources such as platform settlement reports and collective agreements. However, because the skill does not define or use any tools (e.g., shell access, file writing, or network communication), these ingestion points do not pose an exploitation risk.
  • [OBFUSCATION]: No obfuscated content, Base64 encoding, homoglyphs, or hidden URLs were found. The Swedish-language terminology is appropriate for the context and uses standard character sets.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 15, 2026, 09:54 AM