Skip to main content

​
Installation

Memorose runs as a Rust API server plus a separate dashboard UI.

​
Prerequisites

  • Rust toolchain
  • Node.js and pnpm for the dashboard
  • An LLM provider key for embeddings and consolidation
  • Optional: jq for command-line examples

​
Clone the repository

git clone https://github.com/ai-akashic/Memorose.git
cd Memorose

​
Environment setup

Copy the example environment file: 
cp .env.example .env
At minimum, set a provider key plus model names: 
GOOGLE_API_KEY=your_google_api_key_here
LLM_MODEL=gemini-2.0-flash
EMBEDDING_MODEL=text-embedding-004
If you want to force the provider through env overrides, prefer: 
MEMOROSE__LLM__PROVIDER=Gemini
Or: 
MEMOROSE__LLM__PROVIDER=OpenAI
OPENAI_API_KEY=your_openai_api_key_here
To avoid the default dashboard password warning, also set: 
DASHBOARD_ADMIN_PASSWORD=change-me

​
Configuration

Create config.toml yourself using the current schema from crates/memorose-common/src/config.rs, or start from config.example.toml / config.toml.example and keep config.rs as the final source of truth. A minimal starting point is: 
[llm]
provider = "Gemini"
model = "gemini-2.0-flash"
embedding_model = "text-embedding-004"

[storage]
root_dir = "./data/node-1"

[raft]
node_id = 1
raft_addr = "127.0.0.1:5001"
heartbeat_interval_ms = 500
election_timeout_min_ms = 1500
election_timeout_max_ms = 3000
snapshot_logs = 1000000
auto_initialize = true

[worker]
llm_concurrency = 5
decay_interval_secs = 60
prune_threshold = 0.1
consolidation_interval_ms = 1000
community_interval_ms = 1000
insight_interval_ms = 30000
enable_auto_planner = true
enable_task_reflection = true
auto_link_similarity_threshold = 0.6
tick_interval_ms = 100

[reranker]
type = "weighted"
To enable internal model-based arbitration, add a reranker block: 
[reranker]
type = "arbitrator"
model = "gemini-3.1-flash-lite-preview"
max_candidates = 32
fallback_to_weighted = true

​
Start the stack

Recommended local startup: 
./scripts/start_cluster.sh start --clean --build
This starts backend nodes on 3000, 3001, and 3002 in local cluster mode, plus the dashboard UI on 3100. If you want a single local node: 
./scripts/start_cluster.sh start --mode standalone

​
Manual server startup

If you only want the API server: 
cargo run --release -p memorose-server

​
Docker reranker configuration

The Docker image starts the API server on 3000 and the dashboard on 3100. Reranker mode is still configured by the same MEMOROSE__* environment variables: 
docker run -d \
  --name memorose \
  -p 3000:3000 \
  -p 3100:3100 \
  -v memorose_data:/app/data \
  -e GOOGLE_API_KEY="your_google_api_key_here" \
  -e MEMOROSE__LLM__MODEL="gemini-3.1-flash-lite-preview" \
  -e MEMOROSE__LLM__EMBEDDING_MODEL="gemini-embedding-2-preview" \
  -e MEMOROSE__RERANKER__TYPE="arbitrator" \
  -e MEMOROSE__RERANKER__MODEL="gemini-3.1-flash-lite-preview" \
  -e MEMOROSE__RERANKER__MAX_CANDIDATES="32" \
  -e MEMOROSE__RERANKER__FALLBACK_TO_WEIGHTED="true" \
  dylan2024/memorose:latest

​
Verify the installation

Check the root endpoint: 
curl http://127.0.0.1:3000/
Expected response: 
Memorose is running.

​
Dashboard

  • Dashboard UI: http://127.0.0.1:3100/dashboard
  • API redirect: http://127.0.0.1:3000/dashboard
  • Default login: admin / admin
On first login, the response may include must_change_password: true until you set a new password.
If you deploy the dashboard using your own docker-compose.yml or docker run, you must set DASHBOARD_API_ORIGIN on the dashboard container to point at the backend API, for example DASHBOARD_API_ORIGIN=http://memorose-node-0:3000. Otherwise login and proxied dashboard requests will fail with connect ECONNREFUSED 127.0.0.1:3000.