mcporter

Warn

Audited by Gen Agent Trust Hub on Sep 8, 2026

Risk Level: MEDIUMCOMMAND_EXECUTIONCREDENTIALS_UNSAFEEXTERNAL_DOWNLOADSDYNAMIC_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill uses the mcporter CLI which accepts arbitrary shell commands via the --stdio argument (e.g., mcporter list --stdio "npx -y ..."). This allows the agent to execute arbitrary local code or scripts.
  • [CREDENTIALS_UNSAFE]: The skill provides commands for authentication management (mcporter auth) and configuration handling (mcporter config import ), which can be used to access or manipulate sensitive session tokens and credentials.
  • [EXTERNAL_DOWNLOADS]: The skill fetches the mcporter package from the NPM registry and connects to external MCP servers via HTTP URLs (e.g., https://some-mcp-server.com).
  • [DYNAMIC_EXECUTION]: The skill includes functionality to generate executable CLI wrappers (generate-cli) and TypeScript clients (emit-ts) at runtime.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes potentially untrusted data from external MCP tools. * Ingestion points: mcporter call output and tool schemas from mcporter list definitions from remote servers. * Boundary markers: None present; tool outputs are processed directly by the agent. * Capability inventory: Includes shell command execution, file writing, and network operations. * Sanitization: No validation or escaping of tool-returned data is specified.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Sep 8, 2026, 06:54 AM
Security Audit — agent-trust-hub — mcporter