- start-llama-server.bat: fix LLAMA_DIR to bin\llama (was bin\ollama),
fix model to qwen2.5-coder-1.5b (was Llama-3.2-1B), use correct
GPU (llama-server.exe) and CPU (cpu\llama-server-cpu.exe) binaries
- start-windows-mcp.bat: remove dead python313\ tier, use single
embedded Python 3.13 at python\python.exe
- run-hermes.bat: use Scripts\hermes.exe instead of -m hermes_cli
(no __main__.py in package), fix offline mode to use single
config.yaml instead of non-existent config.offline.yaml
- run-hermes.ps1: same hermes_cli → hermes.exe fix
- README.txt: update to match actual USB state (Python 3.13, 142
wheels, Qwen2.5-Coder-1.5B 1117MB model, correct dir structure)
- Fixed quick-rebuild-iso.sh paths from /hermes/ to /opt/hermes/
- Fixed ISO boot structure (uses GRUB eltorito/efi, not isolinux)
- Added MBR extraction fallback for xorriso
- Updated start-llama-server.bat to reference existing USB paths
(bin/ollama/, Llama-3.2-1B model, llama-server-vulkan.exe)
- Works with existing USB structure — no restructure needed
- All bat files: ASCII text, CRLF line terminators
- build/usb-image.sh: Main USB image assembler with artifact validation,
ISO building, disk image creation, and USB write support
- build/ventoy.json: Ventoy configuration for multi-boot USB
- src/hermes/config.yaml: Portable Hermes agent configuration
- src/hermes/autorun.sh: Auto-launch script with full diagnostic pipeline
- Makefile: Build system with iso/image/check/clean/write targets
Generates fallback stubs for any diagnostic modules whose upstream
build tasks haven't completed yet, ensuring the image boots even
during active development.
shawn