alipay-payment-feedback
Pass
Audited by Gen Agent Trust Hub on Sep 19, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes the alipay-bot command-line tool to submit problem feedback. The command uses a --reason flag to pass user-generated content.
- [EXTERNAL_DOWNLOADS]: The skill downloads and installs the @alipay/agent-payment package from the official NPM registry if the alipay-bot CLI is not present. The installation script (install-alipay-cli.sh) implements several security measures, including pinning the version to 1.0.0, verifying the dist.integrity SHA-512 hash, validating the maintainer identity, and performing a signature audit.
- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted user input from the conversation context to populate the problem description field for the feedback command. Ingestion points: Problem descriptions are collected from the conversation context in SKILL.md. Boundary markers: The instructions mandate wrapping the input for the --reason parameter in double quotes. Capability inventory: The skill has the capability to execute the alipay-bot shell command. Sanitization: The skill instructs the agent to organize the problem description before submission and avoid collecting sensitive information.
Audit Metadata