ai-social-media-content

Pass

Audited by Gen Agent Trust Hub on Aug 8, 2026

Risk Level: SAFE
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill provides instructions to install the belt CLI tool using npx skills add belt-sh/cli and recommends several utility skills from the inference-sh/skills repository. These resources originate from the service provider's infrastructure.
  • [COMMAND_EXECUTION]: The skill utilizes the belt CLI via the Bash tool to perform AI inference tasks, such as belt app run google/veo-3-1-fast. This is the intended core functionality of the skill for generating social media assets.
  • [PROMPT_INJECTION]: The provided shell scripts interpolate user-defined variables like $CONCEPT and $topic directly into model prompt strings. While standard for content generation, this represents a surface for indirect prompt injection if the source data is untrusted.
  • Ingestion points: Shell variables in the batch processing and repurposing templates in SKILL.md.
  • Boundary markers: No explicit delimiters or instructions are used to separate user data from the system prompt in the examples.
  • Capability inventory: The skill uses Bash to interact with various generative models via the belt CLI.
  • Sanitization: No input sanitization or validation is implemented in the provided shell templates.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 8, 2026, 08:29 PM
Security Audit — agent-trust-hub — ai-social-media-content