french-native-copy
Pass
Audited by Gen Agent Trust Hub on Oct 11, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to process external content such as English captions and brand briefs to generate French copy, which creates a potential surface for indirect prompt injection if the source material contains malicious instructions.\n
- Ingestion points: The skill ingests "approved English caption or brief, brand voice, or raw client facts" as defined in SKILL.md.\n
- Boundary markers: The instructions do not specify the use of delimiters or markers to separate untrusted source material from the system instructions.\n
- Capability inventory: The skill's capabilities are limited to text analysis and generation; it does not perform file writes, network operations, or shell command execution.\n
- Sanitization: No explicit sanitization or filtering of input content is mentioned.\n- [SAFE]: The skill incorporates several security and quality best practices, including the use of placeholders for missing sensitive information (like prices) and a mandatory 'ship gate' requiring native human review before publication to prevent the delivery of inaccurate or fabricated information.\n- [SAFE]: References to external domains and contact information are transparently provided as author acknowledgments and do not involve data exfiltration or suspicious network activity.
Audit Metadata