meta-analysis

Fail

Audited by Gen Agent Trust Hub on Feb 15, 2026

Risk Level: HIGHPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [Indirect Prompt Injection] (HIGH): The skill processes untrusted external datasets (effect sizes, study results) while having access to powerful tools like Bash and Write.
  • Ingestion points: External data files or user input containing study results.
  • Capability inventory: Use of Bash, Read, and Write tools.
  • Boundary markers: Absent; no instructions to ignore embedded commands in study data.
  • Sanitization: Absent; no validation logic for external content.
  • [Command Execution] (MEDIUM): The skill requests the Bash tool in its metadata, which increases the impact of successful prompt injection through malicious study data.
Recommendations
  • AI detected serious security threats
Audit Metadata
Risk Level
HIGH
Analyzed
Feb 15, 2026, 12:04 AM