cloud-storage

Pass

Audited by Gen Agent Trust Hub on Sep 1, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADSCOMMAND_EXECUTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill ingests file metadata and directory structures from external cloud storage APIs, creating a surface for potential indirect instructions.\n
  • Ingestion points: Data is retrieved from Baidu and Quark APIs during file listing and status check operations (e.g., Baidu's list and quota methods).\n
  • Boundary markers: The instructions do not define clear delimiters or use boundary markers to distinguish between data and instructions when processing API responses.\n
  • Capability inventory: The skill makes use of network operations via curl, shell execution for CLI tools, and local script execution for token management.\n
  • Sanitization: No validation or sanitization logic is provided for external file names or metadata received from the cloud providers.\n- [EXTERNAL_DOWNLOADS]: The documentation instructs the user or agent to install the quarkpan package from PyPI, which is a well-known and standard package registry for Python.\n- [COMMAND_EXECUTION]: The skill uses various shell commands and CLI tools (curl, python3, quarkpan) to perform DevOps tasks such as managing cloud storage authentication, manipulating files, and refreshing access tokens.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 1, 2026, 05:58 AM