galloway-youtube-outlier-packaging
Pass
Audited by Gen Agent Trust Hub on Oct 8, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data, such as channel briefs, niche information, and external video research links, which could contain malicious instructions designed to influence the agent's behavior.
- Ingestion points: Data intake occurs in
SKILL.md(Step 1) and is further defined inreferences/method.md(Section 1). - Boundary markers: While the skill utilizes structured output templates in
templates/output-template.md, it lacks explicit instructions for the agent to ignore or delimit potentially malicious instructions within the external data it processes. - Capability inventory: The skill is restricted to strategic planning and text generation; no evidence of subprocess execution, file writing, or network exfiltration was found.
- Sanitization: There are no instructions for sanitizing or escaping the content retrieved from external sources before it is processed by the agent.
Audit Metadata