li-short-video-eight-step-review

Pass

Audited by Gen Agent Trust Hub on Aug 18, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill consists entirely of instructional markdown content and a reference case study.
  • No executable code, shell commands, or external binary dependencies are present.
  • No network connectivity or sensitive file access is requested.
  • [INDIRECT_PROMPT_INJECTION]: The skill analyzes user-provided scripts, which constitutes a potential injection surface.
  • Ingestion points: Processes user-provided broadcast scripts and VLOG narrations as described in SKILL.md.
  • Boundary markers: No explicit delimiters are mandated for user input, but the instructions direct the AI to perform a diagnostic review rather than follow commands embedded in the input.
  • Capability inventory: There are no active tools or capabilities (e.g., file-write, network, subprocess) available to the skill that could be misused by an injection.
  • Sanitization: Input sanitization is not specified, but the risk is minimal given the lack of exploitable capabilities.
  • [PROMPT_INJECTION]: No instructions were found that attempt to bypass AI safety guardrails, override core agent behavior, or extract system prompts.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 18, 2026, 02:31 AM
Security Audit — agent-trust-hub — li-short-video-eight-step-review