mcp-security-hub

Warn

Audited by Gen Agent Trust Hub on May 18, 2026

Risk Level: MEDIUMEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONDATA_EXFILTRATION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill instructs users to clone a repository containing 38 MCP servers from an external GitHub account (github.com/FuzzingLabs/mcp-security-hub.git).
  • [COMMAND_EXECUTION]: The configuration requires the AI to execute the docker command with arguments that include building and running numerous containers on the local system.
  • [COMMAND_EXECUTION]: The trivy server configuration mounts /var/run/docker.sock, which grants the containerized tool root-level control over the host machine's Docker daemon, representing a high risk for privilege escalation.
  • [COMMAND_EXECUTION]: Several server configurations (e.g., gitleaks, radare2) mount sensitive local paths such as ${HOME}/repos and ${HOME}/binaries into containers, exposing local user data to the AI agent.
  • [DATA_EXFILTRATION]: The skill facilitates the use of network-active tools like Nmap, Nuclei, and SQLMap, which are designed to transmit data to and from external network targets as part of security scanning workflows.
Audit Metadata
Risk Level
MEDIUM
Analyzed
May 18, 2026, 08:58 PM
Security Audit — agent-trust-hub — mcp-security-hub