press-release-writer
Pass
Audited by Gen Agent Trust Hub on May 5, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill implements a security-positive 'no-fly list' check using the
get_marketing_referencetool to ensure that restricted customer names are never included in the generated output. - [SAFE]: Data access is restricted to domain-specific tools (
list_marketing_references,get_marketing_context,search_content) used to fetch relevant marketing context and internal documentation. - [SAFE]: The skill uses a structured multi-step workflow for drafting and copy-editing text without invoking dangerous system commands or performing unauthorized network operations.
- [SAFE]: Analysis of indirect prompt injection surfaces (Step 1 and Step 2) confirms that untrusted data is used solely for text generation within markdown, with no high-privilege capabilities exposed to the content.
Audit Metadata