Gemma4 llama.cpp Hermes profile is ready.

Profile name:
- gemma4-llamacpp

Wrapper command:
- gemma4-llamacpp chat

Direct launcher:
- bash ~/.timmy/scripts/run_hermes_gemma4_llamacpp.sh

Current expectation:
- Hermes profile points at local llama.cpp on http://localhost:8081/v1
- You still need a llama-server process on 8081 with a real GGUF loaded
- Ideally that GGUF is Gemma 4 when we finish the model side

Current blocker:
- there is no active llama-server on :8081 right now
- Gemma4 is available in Ollama, but you said you want llama.cpp, so this profile is prepared for the llama.cpp path specifically