skills/jmagly/aiwg/aiwg-mcp-server/Gen Agent Trust Hub

aiwg-mcp-server

Warn

Audited by Gen Agent Trust Hub on Jun 19, 2026

Risk Level: MEDIUMCOMMAND_EXECUTIONEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill uses unvalidated placeholders such as <client>, <mcp-server-ref>, and <name> within shell command templates (e.g., aiwg mcp install <client>, aiwg mcp add <mcp-server-ref>). This creates a vulnerability to command injection where malicious input containing shell metacharacters (e.g., ;, &&, |) could execute arbitrary commands on the host system.
  • Ingestion points: Placeholders in the 'Behavior' and 'Examples' sections of SKILL.md receive user-provided strings.
  • Boundary markers: No delimiters or escaping mechanisms are used to separate user input from the command string.
  • Capability inventory: The skill is designed to run several shell-based subcommands using the aiwg CLI.
  • Sanitization: There is no evidence of validation or sanitization for the inputs before they are passed to the shell.
  • [EXTERNAL_DOWNLOADS]: The aiwg mcp add <mcp-server-ref> command allows the agent to register external MCP servers from remote references, such as NPM packages or third-party repositories. This introduces code from external sources into the local environment.
  • [REMOTE_CODE_EXECUTION]: By design, registered MCP servers are intended to be executed by the host platform. The ability to add arbitrary external references creates a mechanism for remote code execution if the sources are untrusted or compromised.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Jun 19, 2026, 12:37 PM
Security Audit — agent-trust-hub — aiwg-mcp-server