feat: add menlo build and menlo Makefile

[Minh141120]

This pull request introduces two major improvements: it adds an automated workflow for syncing with the upstream llama.cpp repository, and it provides a new, comprehensive Makefile for building, signing, notarizing, and packaging the Cortex llamacpp engine. The new Makefile is also added to the menlo/ directory for consistency. These changes streamline both the update process and the build process for the project.

Automated Upstream Sync:

• Adds a new GitHub Actions workflow (.github/workflows/upstream-sync.yml) to automatically sync the master branch with the latest upstream release from ggml-org/llama.cpp, create a pull request to update the dev branch, wait for CI checks, merge if successful, and create a new tag based on the commit count.

Build System Improvements:

• Replaces the old Makefile with a new, detailed build system that supports building the engine, code signing, notarization, and packaging for Windows, Linux, and macOS, with configurable environment variables for each step.
• Adds the new Makefile to the menlo/ directory to support local builds and developer workflows in that subdirectory.