create-voltagent

Warn

Audited by Gen Agent Trust Hub on Feb 17, 2026

Risk Level: MEDIUMEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONCOMMAND_EXECUTIONCREDENTIALS_UNSAFEPROMPT_INJECTION
Full Analysis
  • EXTERNAL_DOWNLOADS (MEDIUM): The skill utilizes npm create voltagent-app@latest and installs multiple packages under the @voltagent/ scope. Since the VoltAgent organization is not on the trusted list, these dependencies are considered unverifiable.
  • REMOTE_CODE_EXECUTION (MEDIUM): The command npm create voltagent-app@latest fetches and runs code from the npm registry. This is a high-privilege action that executes arbitrary code from the package's binary script.
  • COMMAND_EXECUTION (LOW): Multiple shell commands are suggested for project setup, directory creation, and starting local development servers (npm run dev, mkdir, etc.).
  • CREDENTIALS_UNSAFE (LOW): The skill prompts users to enter and store provider API keys (OpenAI, Anthropic, etc.) in a local .env file. While typical for development, it requires the user to manage these secrets securely.
  • PROMPT_INJECTION (LOW): Category 8 (Indirect Prompt Injection) vulnerability surface detected: 1. Ingestion points: location parameter in weatherTool and input object in expenseApprovalWorkflow. 2. Boundary markers: Absent in the provided templates. 3. Capability inventory: Local tool execution and workflow orchestration. 4. Sanitization: Uses Zod for schema validation and structured data parsing.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Feb 17, 2026, 05:36 PM