Data Analysis

Pass

Audited by Gen Agent Trust Hub on May 8, 2026

Risk Level: SAFEDATA_EXFILTRATIONPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [DATA_EXFILTRATION]: The skill uses the asta analyze-data submit command to upload local datasets to a remote service hosted on Cloud Run (asta-gateway). This is the intended primary function of the skill for performing cloud-based data analysis.
  • [PROMPT_INJECTION]: The skill processes external data files and user-defined questions, creating a surface for indirect prompt injection (Category 8).
  • Ingestion points: Local dataset files (e.g., CSVs) and confirmation queries provided by the user in Step 4.
  • Boundary markers: The confirmed analytical question is enclosed in double quotes within the shell command; however, there are no specific markers or instructions to ignore potential commands within the data files themselves.
  • Capability inventory: The skill utilizes file reading for uploads, network access via the asta CLI, and local filesystem access via the open command.
  • Sanitization: No explicit sanitization or validation of the content within the uploaded data files is described in the workflow.
  • [COMMAND_EXECUTION]: The skill relies on the execution of the asta suite of tools and the open command to perform its tasks, including authentication, data submission, polling for results, and opening local artifact directories for the user.
Audit Metadata
Risk Level
SAFE
Analyzed
May 8, 2026, 12:04 AM
Security Audit — agent-trust-hub — Data Analysis