talking-head-production
Pass
Audited by Gen Agent Trust Hub on Sep 2, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill provides numerous instructions to execute shell commands using the
beltCLI tool to interact with remote AI inference services for video production. - [EXTERNAL_DOWNLOADS]: The skill references external resources for setup, including an installation skill
npx skills add belt-sh/cliand a remote installation guide located athttps://raw.githubusercontent.com/inference-sh/skills/refs/heads/main/cli-install.md. - [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process data from external URLs, which represents an attack surface for indirect prompt injection.
- Ingestion points: The skill takes several untrusted inputs via the CLI, including
image,audio,image_url,audio_url, andvideo_urlparameters withinSKILL.md. - Boundary markers: No protective delimiters or instructions to ignore embedded commands are present in the documentation for handling these external inputs.
- Capability inventory: The skill utilizes the
Bashtool to run thebeltCLI, which performs network operations and processes media files. - Sanitization: There is no evidence of validation or sanitization for the remote content before it is passed to the AI models.
Audit Metadata