Skills
Navigate
skills/reason-machines/data-skills/realtime-cinema-data-engineering-pipeline/Gen Agent Trust Hub

realtime-cinema-data-engineering-pipeline

Warn

Audited by Gen Agent Trust Hub on Oct 1, 2026

Risk Level: MEDIUMCREDENTIALS_UNSAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [CREDENTIALS_UNSAFE]: The skill contains hardcoded default credentials for the local PostgreSQL database.
  • Evidence: POSTGRES_USER=postgres, POSTGRES_PASSWORD=postgres, and AIRFLOW_CONN_POSTGRES_CINEMA_DW=postgresql://postgres:postgres@postgres:5432/cinema_dw in the environment configuration section.
  • [EXTERNAL_DOWNLOADS]: The skill instructs the agent to clone code from an untrusted repository.
  • Evidence: git clone https://github.com/BaidaneAyoub/realtime-cinema-data-engineering.git in the Setup Steps.
  • [COMMAND_EXECUTION]: The instructions include commands to install and execute the downloaded external code.
  • Evidence: pip install -r requirements.txt, python producer/main_producer.py, python consumer/main_consumer.py, and streamlit run dashboard/app.py.
  • [INDIRECT_PROMPT_INJECTION]: The pipeline ingests untrusted data from a Kafka stream which is then processed by an Airflow ELT pipeline.
    1. Ingestion points: consumer/main_consumer.py reads events from the cinema_transactions Kafka topic.
    1. Boundary markers: None identified; data is parsed as raw JSON.
    1. Capability inventory: The skill performs database writes to the Bronze and Silver layers and executes transformations via Airflow PythonOperators.
    1. Sanitization: No explicit sanitization or validation of the ingested JSON payload structure before processing.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Oct 1, 2026, 01:35 PM
Security Audit — agent-trust-hub — realtime-cinema-data-engineering-pipeline