8. Configuration
All configuration lives in config/config.yaml. Key sections:
api:
base_url: "" # llama.cpp server URL
embedding_model: "" # embedding model name
timeout_seconds: 300 # request timeout
memory:
persist_dir: "data/memory"
embedding_dimension: 768
auto_embed: true
identity:
user_name: ""
assistant_name: "Memo"
style: "default"
system_role: ""
proactive:
enabled: false
level: "off" # off | low | medium | high
calendar:
enabled: false
reminder_lead_minutes: 15
disable_time_guess: false
sync:
enabled: false
passphrase: "" # empty = machine-derived key
interval_messages: 50
learning:
enabled: false