terms-of-service
Pass
Audited by Gen Agent Trust Hub on Oct 2, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill instructions emphasize safety and accuracy, explicitly prohibiting the agent from inventing legal specifics, copying other companies' terms, or claiming legal compliance. It requires marking legal interpretations as 'REVIEW REQUIRED'.
- [EXTERNAL_DOWNLOADS]: The skill provides a detailed reference list of official sources for legal research (e.g., EUR-Lex, FTC, W3C). These are well-known, high-trust domains used for factual reference, and no automated execution of downloaded content is performed.
- [INDIRECT_PROMPT_INJECTION]: The skill processes external data regarding product features to generate terms. It mitigates potential risks by using a structured evidence-labeling system (OBSERVED, SOURCE-INDICATED, DECLARED) and restricting the agent from generating unsupported clauses.
- [DATA_EXPOSURE_EXFILTRATION]: The skill does not access sensitive local files or environment variables. Its focus is on publicly visible product features and official legal documentation.
Audit Metadata