NYC
skills/smithery/ai/jiap-analyst/Gen Agent Trust Hub

jiap-analyst

Fail

Audited by Gen Agent Trust Hub on Feb 17, 2026

Risk Level: CRITICALCOMMAND_EXECUTION
Full Analysis
  • COMMAND_EXECUTION (SAFE): The skill executes local Python scripts (scripts/jadx_check.py, scripts/jadx_analyze.py) to verify the environment and perform decompilation of target APK files. This is the primary intended function and is restricted to local file operations.
  • INDIRECT_PROMPT_INJECTION (LOW): As a security analysis tool, it ingests untrusted third-party data (APKs). 1. Ingestion points: APK files loaded via jadx_analyze.py. 2. Boundary markers: None provided in instructions. 3. Capability inventory: Provides tools for searching and retrieving decompiled source code. 4. Sanitization: No explicit sanitization of untrusted code strings is defined. This represents an inherent but expected surface for this type of tool.
  • SAFE (INFO): The malicious URL detections provided in the prompt are false positives. The string 'com.app/.ExportedActivity' is an Android component reference used in a documentation template for vulnerability reporting, not a network-accessible phishing link.
Recommendations
  • Contains 2 malicious URL(s) - DO NOT USE
Audit Metadata
Risk Level
CRITICAL
Analyzed
Feb 17, 2026, 06:50 PM