interpreting-electrocardiograms

Pass

Audited by Gen Agent Trust Hub on Aug 6, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it requires the agent to ingest and process untrusted external data as part of its primary function.
  • Ingestion points: The 'Documents to Request' section in SKILL.md identifies several external data sources: ECG tracings (PDF/image), prior ECGs, medication lists, BMP reports, clinical notes, echocardiogram reports, and device interrogation reports.
  • Boundary markers: The instructions do not define boundary markers or explicit prompts to ignore instructions embedded within these external documents.
  • Capability inventory: The skill does not contain executable scripts, subprocess calls, or network operations, which significantly limits the potential impact of an injection attack.
  • Sanitization: No sanitization, escaping, or content validation mechanisms are implemented for the external documents provided by the user.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 6, 2026, 07:43 AM
Security Audit — agent-trust-hub — interpreting-electrocardiograms