ai-marketing-videos

Fail

Audited by Gen Agent Trust Hub on Feb 16, 2026

Risk Level: CRITICALREMOTE_CODE_EXECUTIONEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • REMOTE_CODE_EXECUTION (CRITICAL): Found curl -fsSL https://cli.inference.sh | sh. This pattern downloads and executes a script from an untrusted domain directly in the host shell without verification, allowing for immediate system compromise.
  • EXTERNAL_DOWNLOADS (HIGH): The skill utilizes npx skills add to install and execute additional code from inferencesh/skills. As this source is not in the trusted scope, it represents an unverifiable dependency risk.
  • PROMPT_INJECTION (HIGH): The skill is vulnerable to Indirect Prompt Injection (Category 8). Ingestion points: User-provided text prompts for video generation (e.g., in the 'Complete Ad Workflows' section). Boundary markers: Absent; inputs are directly interpolated into JSON-formatted bash arguments. Capability inventory: Command execution via the infsh CLI using Bash(infsh *) permissions. Sanitization: Absent. A malicious prompt could escape the shell command or the JSON structure to execute arbitrary code.
  • COMMAND_EXECUTION (HIGH): The skill extensively uses shell commands where user-controlled prompts are interpolated into double-quoted strings (e.g., infsh ... --input "{...$section...}"). This lack of sanitization and reliance on high-privilege shell access provides a significant attack vector.
Recommendations
  • CRITICAL: Downloads and executes remote code from untrusted source(s): https://cli.inference.sh - DO NOT USE
  • AI detected serious security threats
Audit Metadata
Risk Level
CRITICAL
Analyzed
Feb 16, 2026, 06:18 AM