report

Pass

Audited by Gen Agent Trust Hub on Jul 1, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
  • [SAFE]: The skill accesses medical data stored within the data/ directory, including biochemical tests and medication records. This file access is restricted to the skill's operational scope and does not involve sensitive system credentials or configuration files.
  • [EXTERNAL_DOWNLOADS]: The generated HTML reports utilize standard web libraries (Tailwind CSS, Chart.js, and Lucide Icons) hosted on well-known CDNs (cdn.tailwindcss.com, cdn.jsdelivr.net, and unpkg.com). These resources are used for styling and data visualization.
  • [PROMPT_INJECTION]: The skill processes untrusted data from multiple medical record files and interpolates it into a generated HTML template. This creates a surface for Indirect Prompt Injection, specifically Cross-Site Scripting (XSS), where malicious content in the source data could execute in a browser environment. Ingestion points: data/profile.json, data/生化检查/*.json, data/影像检查/*.json, data/medications/medications.json, data/radiation-records.json, data/allergies.json. Boundary markers: Absent. Capability inventory: Read and Write tools across all skill files. Sanitization: No explicit sanitization or validation of the input data is described in the core flow.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 1, 2026, 05:26 AM
Security Audit — agent-trust-hub — report