cloudbase

Pass

Audited by Gen Agent Trust Hub on Nov 17, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSDYNAMIC_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill downloads official CloudBase plugins and skill packages from the vendor's GitHub repositories (e.g., TencentCloudBase/cloudbase-plugin and TencentCloudBase/cloudbase-skills) to set up the development environment.
  • [DYNAMIC_EXECUTION]: During the site onboarding process, the skill uses inline Node.js commands (node -e) to programmatically create or update the .cloudbase/project.json configuration file based on user environment selection.
  • [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest local project metadata (such as project.json, app.json, and README.md) to determine deployment status and project type. This ingestion surface is a standard requirement for the skill's primary function of cloud resource management.
  • Ingestion points: .cloudbase/project.json, project.config.json, app.json, README.md
  • Boundary markers: Not explicitly utilized for data parsing, but tool execution is constrained by user-authorized environment bindings.
  • Capability inventory: File system writes (fs.writeFileSync), shell command execution (tcb, npx), and network operations via official API endpoints.
  • Sanitization: Uses standard JSON serialization for writing configuration files.
Audit Metadata
Risk Level
SAFE
Analyzed
Nov 17, 2026, 06:08 AM