http-api-cloudbase

Pass

Audited by Gen Agent Trust Hub on Mar 9, 2026

Risk Level: SAFE
Full Analysis
  • [PROMPT_INJECTION]: The skill uses instructional language to guide agent behavior for API interaction. No safety bypasses, behavioral overrides, or malicious jailbreak attempts were detected.
  • [CREDENTIALS_UNSAFE]: Authentication tokens and API keys are discussed as requirements. The skill provides clear warnings about security and uses placeholders for examples. No sensitive hardcoded credentials were found.
  • [DATA_EXFILTRATION]: The skill instructs the agent to communicate with official Tencent CloudBase API gateways. These network operations are consistent with the vendor's intended functionality.
  • [EXTERNAL_DOWNLOADS]: The skill utilizes a built-in tool to retrieve OpenAPI specifications. This is a safe and controlled method for accessing documentation.
  • [COMMAND_EXECUTION]: Shell commands provided are limited to standard curl requests for API testing. No dangerous patterns such as shell piping or arbitrary command execution are present.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 9, 2026, 03:49 AM