NYC

cloud-storage-web

Fail

Audited by Gen Agent Trust Hub on Feb 15, 2026

Risk Level: HIGHEXTERNAL_DOWNLOADSPROMPT_INJECTIONDATA_EXFILTRATION
Full Analysis
  • EXTERNAL_DOWNLOADS (HIGH): Automated security scanners identified the malicious URL file.co associated with this skill's content. Blacklisted domains pose a severe risk of malware delivery or malicious script execution.
  • PROMPT_INJECTION (HIGH): The skill is highly vulnerable to Indirect Prompt Injection.
  • Ingestion points: The skill processes filePath via fileInput.files[0], which constitutes untrusted external content provided by the user/browser.
  • Boundary markers: Absent. The skill provides no instructions for the agent to use delimiters or to ignore instructions that might be embedded in file metadata (like filenames) or file content.
  • Capability inventory: The skill provides direct implementation for uploadFile, deleteFile, and getTempFileURL, giving the agent significant write and management access to cloud storage.
  • Sanitization: While the 'Best Practices' section mentions validating file types, the provided code snippets do not implement any sanitization or validation logic, leaving the implementation vulnerable to manipulation.
  • DATA_EXFILTRATION (MEDIUM): The functionality to generate temporary download URLs (getTempFileURL) could be exploited to exfiltrate private cloud data to unauthorized parties if the agent's logic is subverted through the injection surface.
  • EXTERNAL_DOWNLOADS (MEDIUM): The skill depends on the @cloudbase/js-sdk package. Since this package and its organization (Tencent) are not within the defined 'Trusted GitHub Organizations', it is classified as an unverifiable dependency.
Recommendations
  • AI detected serious security threats
  • Contains 1 malicious URL(s) - DO NOT USE
Audit Metadata
Risk Level
HIGH
Analyzed
Feb 15, 2026, 11:36 PM