content-growth-studio
Pass
Audited by Gen Agent Trust Hub on Sep 25, 2026
Risk Level: SAFEOBFUSCATIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [OBFUSCATION]: The file
agents/openai.yamlcontains a Byte Order Mark (BOM, U+FEFF) at the start of the file. While often an encoding artifact from certain text editors, zero-width characters can be used to bypass simple text filters. - [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data from URLs, transcripts, and user notes to generate content, creating a potential vector for indirect prompt injection.
- Ingestion points: The
source_artifactsfield inSKILL.mdandreferences/02-input-contracts-and-gates.mdaccepts external content. - Boundary markers: The instructions lack explicit delimiters to separate untrusted source text from the agent's core instructions.
- Capability inventory: Capabilities are limited to content analysis and production; no dangerous system commands or network operations were identified.
- Sanitization: There is no evidence of sanitization or filtering of the ingested external content.
Audit Metadata