alibabacloud-executing-plans

Pass

Audited by Gen Agent Trust Hub on Jul 12, 2026

Risk Level: SAFE
Full Analysis
  • [COMMAND_EXECUTION]: Uses local bash commands to search for internal guide files and execute Python helper scripts for data processing.
  • [REMOTE_CODE_EXECUTION]: Executes infrastructure management commands (plan, apply, destroy) via the Alibaba Cloud IaC Service using a specialized MCP tool (AlibabaCloud___CallCLI). These operations are restricted to the vendor's cloud management API.
  • [DYNAMIC_EXECUTION]: Generates a self-contained HTML file (topology.html) containing JavaScript for SVG rendering and executes Python snippets to filter and aggregate infrastructure state information.
  • [DATA_EXPOSURE]: Reads local Terraform files (.tf) and task metadata to provide necessary context for remote infrastructure deployment. The skill includes a 'slimming' logic to ensure sensitive data like passwords or ARNs are excluded from generated visualizations.
  • [PROMPT_INJECTION]: Ingests remote data from the IaC Service API to generate topology reports. This is handled as an indirect injection surface, mitigated by structured parsing and attribute filtering.
  • Ingestion points: aliyun iacservice get-execute-state API response.
  • Boundary markers: None explicitly defined for the remote JSON payload.
  • Capability inventory: Remote command execution (AlibabaCloud___CallCLI), local command execution (bash), and local file writes (Write).
  • Sanitization: Employs a Python script with a strict allow-list of over 50 specific keys (e.g., vpc_id, instance_type) to filter incoming state data before use.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 12, 2026, 01:50 PM
Security Audit — agent-trust-hub — alibabacloud-executing-plans