EN
Configuration Reference
Memo is configured via config.yaml. All sections and keys are documented below.
api
API server and external provider settings.
| Key | Type | Default | Description |
|---|---|---|---|
base_url |
string | "http://localhost:8080" |
Base URL for the Memo API server |
embedding_model |
string | "all-MiniLM-L6-v2" |
Model used for generating text embeddings |
timeout |
integer | 120 |
Request timeout in seconds |
memory
Persistent memory and vector store settings.
| Key | Type | Default | Description |
|---|---|---|---|
persist_dir |
string | "~/.memo/memory" |
Directory for memory database and indexes |
embedding_dimension |
integer | 384 |
Dimension of embedding vectors (model-dependent) |
auto_embed |
boolean | true |
Automatically embed new memories on creation |
identity
Assistant persona and behavior.
| Key | Type | Default | Description |
|---|---|---|---|
user_name |
string | "User" |
How the assistant addresses you |
assistant_name |
string | "Memo" |
Name of the assistant |
style |
string | "friendly" |
Conversation style (friendly, professional, concise) |
system_role |
string | "" |
Custom system prompt override |
proactive
Proactive suggestion and agent behavior.
| Key | Type | Default | Description |
|---|---|---|---|
enabled |
boolean | false |
Enable proactive suggestions from the assistant |
level |
string | "low" |
Proactivity level (low, medium, high) |
calendar
Calendar integration and reminders.
| Key | Type | Default | Description |
|---|---|---|---|
enabled |
boolean | false |
Enable calendar integration |
reminder_lead_minutes |
integer | 15 |
Minutes before an event to trigger a reminder |
sync
End-to-end encrypted cloud sync settings.
| Key | Type | Default | Description |
|---|---|---|---|
enabled |
boolean | false |
Enable cloud sync between devices |
passphrase |
string | "" |
Encryption passphrase for sync data |
interval_messages |
integer | 100 |
Number of messages between sync checkpoints |
learning
Adaptive learning settings.
| Key | Type | Default | Description |
|---|---|---|---|
enabled |
boolean | true |
Enable the assistant learning from interactions |
llama
Local LLM server configuration.
| Key | Type | Default | Description |
|---|---|---|---|
host |
string | "127.0.0.1" |
Llama server host address |
port |
integer | 0 |
Llama server port (0 = auto-assign) |
model_path |
string | "" |
Path to GGUF model file |
gpu_layers |
integer | -1 |
Layers to offload to GPU (-1 = all) |
context_size |
integer | 4096 |
Context window size in tokens |
threads |
integer | 0 |
CPU threads (0 = auto-detect) |