narrator-ai-cli-video-narration
Pass
Audited by Gen Agent Trust Hub on Jun 17, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: Fetches the primary toolset from the NarratorAI-Studio repository on GitHub as part of the installation process.
- [COMMAND_EXECUTION]: Orchestrates video production by executing a series of CLI commands. Included Python automation examples demonstrate the use of subprocess calls for task polling and resource management.
- [PROMPT_INJECTION]: The skill features a surface for indirect prompt injection by processing external data (movie titles and descriptions) that are interpolated into shell commands.
- Ingestion points: Untrusted movie metadata (e.g., drama_name, drama_intro) enters the context from user input or external searches.
- Boundary markers: No delimiters or instructions are used to separate untrusted data from command arguments.
- Capability inventory: Extensive use of
subprocess.run(shell=True)and CLI commands for file operations and task creation. - Sanitization: The provided script examples do not demonstrate validation or escaping of variables before they are interpolated into shell strings.
Audit Metadata