audio-sync-assembly
Pass
Audited by Gen Agent Trust Hub on Jun 14, 2026
Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [PROMPT_INJECTION]: The skill exhibits an attack surface for indirect prompt injection by processing external data from a timeline manifest.
- Ingestion points: Data is ingested from
templates/timeline-manifest.example.jsoncontaining file paths and durations. - Boundary markers: No specific delimiters or safety instructions are defined for the JSON data ingestion.
- Capability inventory: The skill uses
ffprobeand generatesffmpegcommand lines based on the manifest data. - Sanitization: No sanitization or validation logic is specified for the fields read from the manifest.
- [COMMAND_EXECUTION]: The skill instructs the agent to build and execute shell commands for video processing. While these are legitimate tools for the task, the dynamic assembly of commands from user-provided manifest data poses a potential command injection risk if filenames or parameters are not properly escaped.
Audit Metadata