airflow-translations

Pass

Audited by Gen Agent Trust Hub on Jul 31, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill instructs the agent to execute the command 'prek run --from-ref main --hook-stage pre-commit'. This appears to be a typographical error for the standard 'pre-commit' utility, but referencing unknown command names is a security best practice violation that could lead to unintended binary execution. \n- [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection (Category 8) as it ingests and processes untrusted translation content from multiple JSON files. \n
  • Ingestion points: The agent reads and modifies JSON translation files within the locales directory (e.g., 'airflow-core/src/airflow/ui/public/i18n/locales/'). \n
  • Boundary markers: No explicit delimiters or instructions are provided to the agent to treat the content of these files as data rather than instructions. \n
  • Capability inventory: The skill utilizes shell command execution ('breeze', 'prek') and file system modifications ('mkdir'). \n
  • Sanitization: No mechanism for sanitizing or validating the translation strings is mentioned before they are processed by development tools.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 31, 2026, 08:43 PM
Security Audit — agent-trust-hub — airflow-translations