soia-pkm-alipan-drive-ops

Pass

Audited by Gen Agent Trust Hub on Jul 21, 2026

Risk Level: SAFE
Full Analysis
  • [COMMAND_EXECUTION]: The skill interacts with the aliyunpan command-line tool using wrapper scripts (run_with_env.py and scan_drive.py). These scripts utilize subprocess.run with shell=False and enforce a whitelist of allowed binaries (aliyunpan, aliyunpan.exe), which effectively prevents arbitrary shell command injection.
  • [CREDENTIALS_UNSAFE]: Security best practices are followed for credential management. The skill uses a dedicated script (alipan_env.py) to load sensitive environment variables from a private configuration file. Furthermore, the instructions explicitly forbid the agent from printing or leaking tokens, cookies, or session secrets in its output or logs.
  • [EXTERNAL_DOWNLOADS]: The skill requires the aliyunpan CLI, a well-known tool for Aliyun Drive. The instructions guide the user to install it via standard package managers like Homebrew, ensuring a trusted source for the dependency.
  • [SAFE]: The overall architecture of the skill emphasizes safety, including user-in-the-loop verification for destructive actions like delete or move, and a clear "ops-playbook" for handling sensitive operations like remote login recovery without exposing user credentials.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 21, 2026, 05:46 AM
Security Audit — agent-trust-hub — soia-pkm-alipan-drive-ops