vercel-automation

Pass

Audited by Gen Agent Trust Hub on Mar 29, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill requires the configuration of an external MCP server located at https://rube.app/mcp. This third-party service acts as an intermediary for Vercel platform operations and authentication.\n- [PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface because it grants the agent high-privilege capabilities to modify platform configurations without providing safety constraints for handling untrusted data.\n
  • Ingestion points: The agent may ingest untrusted data from project files, repository content, or external inputs during the automation of Vercel tasks.\n
  • Boundary markers: There are no boundary markers or instructions present to help the agent distinguish between user-intended commands and malicious instructions embedded in processed data.\n
  • Capability inventory: The skill provides tools for high-impact actions including VERCEL_ADD_ENVIRONMENT_VARIABLE, VERCEL_CREATE_DNS_RECORD, VERCEL_DELETE_ENVIRONMENT_VARIABLE, and VERCEL_CREATE_NEW_DEPLOYMENT.\n
  • Sanitization: No sanitization, validation, or escaping mechanisms are described to prevent the injection of malicious payloads into platform configurations.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 29, 2026, 03:31 PM