omnicaptions-translate

Pass

Audited by Gen Agent Trust Hub on Aug 3, 2026

Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [PROMPT_INJECTION]: The skill processes user-supplied subtitle files (SRT, VTT, ASS), which are untrusted external data sources. This creates a potential surface for indirect prompt injection where instructions hidden within the subtitle text could attempt to hijack the agent's behavior during translation.\n
  • Ingestion points: The input_file parameter allows the agent to read and process external subtitle files.\n
  • Boundary markers: The instructions do not specify the use of clear delimiters or warnings for the model to ignore instructions embedded in the subtitle content.\n
  • Capability inventory: The skill uses the agent's native language capabilities and the omnicaptions CLI for processing.\n
  • Sanitization: No explicit sanitization or filtering of the subtitle content is defined.\n- [COMMAND_EXECUTION]: The skill instructs the agent to use the omnicaptions CLI tool for translation. This involves executing shell commands constructed with user-provided parameters like filenames and language codes, which requires proper validation by the underlying agent environment to prevent command injection.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 3, 2026, 07:53 PM
Security Audit — agent-trust-hub — omnicaptions-translate