mempalace-cognition

You are now operating with Long-Term Memory capabilities.

The Brain Lifecycle (llama-server)

For semantic (concept) search to function, the local background brain must be running. If a search command prints a warning that the server is down or unreachable:

1. Immediately run the following command in the background (is_background: true): llama-server --model ~/.gemini/models/nomic-embed-text-v1.5.Q4_K_M.gguf --embedding --port 8080 --threads 2 --no-mmap
2. Wait ~5-10 seconds for it to fully load and bind to the port.
3. Re-run your search or add command.

Taxonomy Guidelines (Wings & Rooms)

When using the add command, enforce a strict and predictable taxonomy. Format: python3 ~/.gemini/MemPalace_Manager.py add "<Wing>" "<Room>" "<Content>"