ai-avatar-video
Pass
Audited by Gen Agent Trust Hub on Nov 18, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill references and encourages the installation of the 'belt-sh/cli' package and provides links to external documentation hosted on GitHub ('inference-sh/skills') for CLI setup.
- [INDIRECT_PROMPT_INJECTION]: The skill provides an interface for processing user-supplied data, such as scripts, image URLs, and audio URLs, which are interpolated into shell commands.
- Ingestion points: User input is intended to be embedded in JSON strings passed to the 'belt' CLI tool via the '--input' flag, as seen in the 'SKILL.md' examples.
- Boundary markers: The instructions do not specify the use of delimiters or 'ignore embedded instructions' warnings when the agent processes this untrusted external content.
- Capability inventory: The skill leverages the 'Bash' tool to execute 'belt' commands that interact with remote APIs ('inference.sh').
- Sanitization: No explicit sanitization or validation logic is provided to handle potentially malicious or malformed user-supplied strings before they are executed.
Audit Metadata