maishou

Pass

Audited by Gen Agent Trust Hub on May 18, 2026

Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTIONDATA_EXFILTRATION
Full Analysis
  • [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection as it retrieves and processes untrusted product information from external APIs.
  • Ingestion points: scripts/main.py (via aiohttp responses from msapi.maishou88.com).
  • Boundary markers: Absent; API data is formatted as CSV/YAML and returned directly to the agent.
  • Capability inventory: Subprocess execution via 'uv run' and network operations.
  • Sanitization: No sanitization is performed on external data strings before presentation.
  • [COMMAND_EXECUTION]: The SKILL.md file defines shell command templates that interpolate user input ({keyword}, {goodsId}) directly into command line arguments, creating a potential injection surface if not properly handled by the agent platform.
  • [DATA_EXFILTRATION]: The skill performs network operations to non-whitelisted domains (maishou88.com and kuaizhan.com) to fetch product data as its primary function.
Audit Metadata
Risk Level
SAFE
Analyzed
May 18, 2026, 09:35 AM
Security Audit — agent-trust-hub — maishou