cro
Pass
Audited by Gen Agent Trust Hub on Oct 11, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill analyzes external websites and local files, which are ingestion points for potentially untrusted data. This presents an indirect prompt injection attack surface where malicious content on a webpage could attempt to subvert the agent's instructions.\n
- Ingestion points: Processes external URLs and local
.agents/product-marketing.mdfiles.\n - Boundary markers: No explicit delimiters are specified for separating analyzed content from instructions.\n
- Capability inventory: The skill is restricted to generating textual analysis and copy suggestions; it possesses no tools for file modification, network exfiltration, or shell command execution.\n
- Sanitization: No data sanitization or input filtering is mentioned in the instructions.\n- [EXTERNAL_DOWNLOADS]: The instructions reference a marketing playbook on GitHub (
coreyhaines31/marketingskills). This is a standard informational reference to a markdown document on a well-known service and does not constitute a risk for remote code execution or unauthorized dependency loading.
Audit Metadata