ai-video-generation
Pass
Audited by Gen Agent Trust Hub on Apr 16, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill utilizes the
Bashtool to execute theinfshCLI for video generation and utility tasks. The skill implements the principle of least privilege by using theallowed-toolsconfiguration to restrict shell access exclusively to theinfshcommand. - [EXTERNAL_DOWNLOADS]: The skill references external installation instructions and documentation from the vendor's domain (
inference.sh) and GitHub repository (github.com/inference-sh). These downloads are necessary for the CLI tool that the skill is designed to interact with. - [PROMPT_INJECTION]: The skill has an attack surface for indirect prompt injection as it processes user-supplied prompts and media URLs that are passed to external models.
- Ingestion points: User-provided strings and URLs are accepted via the
--inputJSON payload in theinfsh app runcommand examples (SKILL.md). - Boundary markers: The skill uses JSON structures to separate data from the CLI command arguments.
- Capability inventory: The skill possesses the ability to execute shell commands (restricted to the
infshbinary) via theBashtool. - Sanitization: There is no evidence of input sanitization or validation within the skill instructions; it relies on the downstream platform to handle potentially malicious payloads.
Audit Metadata