Add policy enforcement for terminal commands and file operations

[L1AD]

This server gives agents direct access to the host operating system:

• execute_command -- runs arbitrary terminal commands
• write_file -- writes to any path on the filesystem
• kill_process -- terminates running processes

execute_command with arbitrary input is effectively root access. An agent can run rm -rf, install packages, modify system files, or exfiltrate data via network commands.

Suggestion: Consider documenting support for PolicyLayer/Intercept, an open-source MCP proxy that enforces policies on tool calls before execution.

Example policy:

version: "1"
default: deny

tools:
  execute_command:
    rules:
      - action: require_approval

  write_file:
    rules:
      - rate_limit: 10/minute

  kill_process:
    rules:
      - action: require_approval

hide:
  - execute_command

One line to set up: npx -y @policylayer/intercept init