create-viral-content

Warn

Audited by Gen Agent Trust Hub on Feb 17, 2026

Risk Level: MEDIUMEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONPROMPT_INJECTION
Full Analysis
  • EXTERNAL_DOWNLOADS (MEDIUM): The skill is hosted on an untrusted GitHub repository (github.com/ice-ninja/skills) and references an external dependency ('humanize-writing') that is not from a trusted organization. This poses a potential supply chain risk as the source is not verified.
  • REMOTE_CODE_EXECUTION (MEDIUM): The skill documentation in 'humanize-integration.md' describes dynamic loading of the 'humanize-writing' skill at runtime when referenced. This mechanism executes code/instructions from an untrusted author without prior validation.
  • PROMPT_INJECTION (LOW): The skill is susceptible to Indirect Prompt Injection because it ingests untrusted user-provided 'topics' or 'ideas' for content generation. Evidence Chain: (1) Ingestion point: User-provided topics via input prompts (README.md); (2) Boundary markers: Absent in provided resource files; (3) Capability: Invokes external skills and generates structured social media content (humanize-integration.md); (4) Sanitization: No evidence of validation, escaping, or filtering for the user input.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Feb 17, 2026, 05:41 PM
Security Audit — agent-trust-hub — create-viral-content