cit-prepay-check-free

Pass

Audited by Gen Agent Trust Hub on Sep 28, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The SKILL.md file provides instructions for purchasing a paid version that involves running npx -y @alipay/agent-payment@latest install. This command fetches code from the official NPM registry for the well-known service provider @alipay.
  • [REMOTE_CODE_EXECUTION]: The suggested npx command executes a remote package. This is a legitimate part of the skill's upgrade process targeting a well-known service and is considered safe practice for the described workflow.
  • [COMMAND_EXECUTION]: The skill is designed to be executed via the command line using Node.js (node scripts/run.mjs). This is the intended operational mode for the tool and is clearly documented for the user.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes user-supplied data in scripts/engine/cit-prepay-check.js.
  • Ingestion point: The text property of the input payload.
  • Boundary markers: No explicit boundaries or warnings are used in the parsing code.
  • Capability inventory: The skill environment has Bash and Read tool access, and the local scripts have file system read access.
  • Sanitization: The code uses strict numeric coercion via normNumber and rateValue functions, which effectively filters non-numeric content from the mathematical calculation engine.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 28, 2026, 09:18 PM