inventory-provision-check-free

Warn

Audited by Gen Agent Trust Hub on Sep 16, 2026

Risk Level: MEDIUMPROMPT_INJECTIONREMOTE_CODE_EXECUTIONEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The SKILL.md file includes a template for users to copy and paste to the AI. This template contains explicit instructions for the AI to bypass user confirmation steps (using the phrase '不要替我确认') when installing dependencies and performing purchases, which overrides the agent's standard safety behavior.\n- [REMOTE_CODE_EXECUTION]: The skill instructs the agent to run npx -y @alipay/agent-payment@latest install. This command downloads and executes code from an external registry. While the source is a well-known service, the practice of executing remote scripts based on instructions in a skill is a significant capability.\n- [EXTERNAL_DOWNLOADS]: The skill points to several external URLs for documentation and transactions, including skillpay.alipay.com and the author's own site at chenqg618.github.io.\n- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted text data (financial tables) via scripts/engine/inventory-provision-check.js.\n
  • Ingestion points: payload.text in scripts/engine/inventory-provision-check.js.\n
  • Boundary markers: Absent.\n
  • Capability inventory: Read and Bash access.\n
  • Sanitization: Input is parsed using regular expressions and converted to numbers, but no explicit instructions are provided to the agent to ignore potential commands hidden in the spreadsheet data.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Sep 16, 2026, 11:28 AM
Security Audit — agent-trust-hub — inventory-provision-check-free