### Harness Setup and Dependency Installation

Source: https://github.com/mystenlabs/memwal/blob/dev/services/server/benchmarks/README.md

Sets up a Python virtual environment, installs necessary benchmark dependencies, and copies the example configuration file. Edit `config.yaml` to set delegate key, account ID, and judge API key.

```bash
cd services/server/benchmarks
python3 -m venv .venv && source .venv/bin/activate
pip install datasets huggingface_hub openai httpx pynacl numpy pyyaml tabulate tqdm pytest
cp config.example.yaml config.yaml
#   edit config.yaml: delegate_key, account_id, judge.api_key (OpenRouter/OpenAI)
```

--------------------------------

### Install Dependencies and Run Development Server

Source: https://github.com/mystenlabs/memwal/blob/dev/apps/noter/README.md

Installs project dependencies, sets up environment variables, generates database tables, and starts the development server.

```bash
pnpm install

cp .env.example .env
# Edit .env with your values

pnpm db:push

pnpm dev
```

--------------------------------

### Install Dependencies and Start Docs Server

Source: https://github.com/mystenlabs/memwal/blob/dev/docs/contributing/run-docs-locally.md

Installs project dependencies and starts the local development server for the documentation site using pnpm. Ensure Node.js 20 LTS is used, as newer versions may cause Mintlify to fail.

```bash
npx -p node@20 -c 'node -v'
pnpm install
pnpm dev:docs
```

--------------------------------

### Install Dependencies, Migrate Database, and Run Development Server

Source: https://github.com/mystenlabs/memwal/blob/dev/apps/chatbot/README.md

Installs project dependencies, applies database migrations, and starts the local development server. Ensure your .env.local file is configured with necessary environment variables.

```bash
pnpm install
pnpm db:migrate # Setup database or apply latest database changes
pnpm dev
```

--------------------------------

### Local Infrastructure Setup

Source: https://github.com/mystenlabs/memwal/blob/dev/services/server/benchmarks/README.md

Starts the necessary Docker containers for Postgres and Redis, which are required for the server to boot.

```bash
docker compose -f services/server/docker-compose.yml up -d
```

--------------------------------

### Benchmark Harness Setup

Source: https://github.com/mystenlabs/memwal/blob/dev/services/server/benchmarks/README.md

Sets up a Python virtual environment, installs necessary dependencies, and configures the benchmark harness. Ensure you have Python 3.9+ installed. Edit `config.yaml` with your delegate key, account ID, and OpenRouter API key.

```bash
cd services/server/benchmarks

# Create venv (Python 3.9+)
python3 -m venv .venv
source .venv/bin/activate

# Install deps directly (don't use pip install -e on Python 3.9)
pip install datasets huggingface_hub openai httpx pynacl numpy pyyaml tabulate tqdm pytest

# Configure
cp config.example.yaml config.yaml
# Edit config.yaml: delegate_key, account_id, OpenRouter API key
```

--------------------------------

### Install memwal SDK

Source: https://github.com/mystenlabs/memwal/blob/dev/docs/python-sdk/quick-start.md

Install the core memwal Python SDK using pip. This is the recommended starting point for most users.

```bash
pip install memwal
```

--------------------------------

### Configure and Run Relayer

Source: https://github.com/mystenlabs/memwal/blob/dev/docs/contributing/run-repo-locally.md

Configure the relayer service by copying the example environment file and editing it with your credentials. Then, install sidecar dependencies and run the relayer using Cargo.

```bash
# Configure environment
cp services/server/.env.example services/server/.env
# Edit .env with your credentials

# Install sidecar dependencies
cd services/server/scripts && npm ci && cd ..

# Run the relayer
cargo run
```

--------------------------------

### Synchronous Quick Start with MemWalSync

Source: https://github.com/mystenlabs/memwal/blob/dev/packages/python-sdk-memwal/README.md

Provides a synchronous example for remembering and recalling data using the MemWalSync client. This version is suitable for applications where asynchronous operations are not required.

```python
import os
from memwal import MemWalSync, RecallParams

client = MemWalSync.create(
    key=os.environ["MEMWAL_PRIVATE_KEY"],
    account_id=os.environ["MEMWAL_ACCOUNT_ID"],
    server_url=os.environ.get("MEMWAL_SERVER_URL", "https://relayer.memory.walrus.xyz"),
)

result = client.remember_and_wait("I'm allergic to peanuts")
matches = client.recall(RecallParams(query="food allergies"))
client.close()
```

--------------------------------

### Install memwal with all integrations

Source: https://github.com/mystenlabs/memwal/blob/dev/docs/python-sdk/quick-start.md

Install the memwal SDK with all available integrations. This includes LangChain, OpenAI, and any other optional dependencies.

```bash
pip install memwal[all]
```

--------------------------------

### Start Local Services

Source: https://github.com/mystenlabs/memwal/blob/dev/README.md

Start specific services for local development using pnpm.

```bash
pnpm dev:app
pnpm dev:noter
pnpm dev:chatbot
pnpm dev:researcher
```

--------------------------------

### Install MemWal SDK with bun

Source: https://github.com/mystenlabs/memwal/blob/dev/docs/getting-started/quick-start.md

Install the main Walrus Memory SDK package using bun.

```bash
bun add @mysten-incubation/memwal
```

--------------------------------

### Install manual client flow packages with bun

Source: https://github.com/mystenlabs/memwal/blob/dev/docs/getting-started/quick-start.md

Install packages required for the manual client flow using bun.

```bash
bun add @mysten/sui @mysten/seal @mysten/walrus
```

--------------------------------

### Install Repository Dependencies

Source: https://github.com/mystenlabs/memwal/blob/dev/README.md

Install all project dependencies using pnpm.

```bash
pnpm install
```

--------------------------------

### Start Server in Benchmark Mode

Source: https://github.com/mystenlabs/memwal/blob/dev/services/server/benchmarks/README.md

Starts the server after configuring it for benchmark mode. Wait for the confirmation log message indicating that `BENCHMARK_MODE=true` is active and the `PlaintextEngine` is being used.

```bash
cd services/server && cargo run
```

--------------------------------

### Install manual client flow packages with npm

Source: https://github.com/mystenlabs/memwal/blob/dev/docs/getting-started/quick-start.md

Install packages required for the manual client flow using npm.

```bash
npm install @mysten/sui @mysten/seal @mysten/walrus
```

--------------------------------

### Install Walrus Memory SDK

Source: https://github.com/mystenlabs/memwal/blob/dev/README.md

Install the main Walrus Memory package using pnpm.

```bash
pnpm add @mysten-incubation/memwal
```

--------------------------------

### Install manual client flow packages with yarn

Source: https://github.com/mystenlabs/memwal/blob/dev/docs/getting-started/quick-start.md

Install packages required for the manual client flow using yarn.

```bash
yarn add @mysten/sui @mysten/seal @mysten/walrus
```

--------------------------------

### Install MemWal SDK with npm

Source: https://github.com/mystenlabs/memwal/blob/dev/docs/getting-started/quick-start.md

Install the main Walrus Memory SDK package using npm.

```bash
npm install @mysten-incubation/memwal
```

--------------------------------

### Install Vercel AI SDK with bun

Source: https://github.com/mystenlabs/memwal/blob/dev/docs/getting-started/quick-start.md

Install the optional Vercel AI SDK package for AI middleware using bun.

```bash
bun add ai
```

--------------------------------

### Install MemWal SDK with yarn

Source: https://github.com/mystenlabs/memwal/blob/dev/docs/getting-started/quick-start.md

Install the main Walrus Memory SDK package using yarn.

```bash
yarn add @mysten-incubation/memwal
```

--------------------------------

### MemWal.create(config)

Source: https://github.com/mystenlabs/memwal/blob/dev/docs/llms-full.txt

Initializes a new MemWal instance with the provided configuration. This is the recommended way to start using the MemWal SDK.

```APIDOC
## MemWal.create(config)

### Description
Initializes a new MemWal instance with the provided configuration. This is the recommended way to start using the MemWal SDK.

### Method
`create(config: MemWalConfig): MemWal`

### Parameters
#### Request Body
- **config** (MemWalConfig) - Required - Configuration object for MemWal.
  - **key** (string) - Required - Ed25519 delegate private key in hex.
  - **accountId** (string) - Required - MemWalAccount object ID on Sui.
  - **serverUrl** (string) - Optional - Relayer URL. Defaults to `https://relayer.memory.walrus.xyz`.
  - **namespace** (string) - Optional - Default namespace for memory isolation. Defaults to `"default"`.
```

--------------------------------

### Install Peer Dependencies

Source: https://github.com/mystenlabs/memwal/blob/dev/README.md

Install necessary peer dependencies for Walrus Memory using pnpm.

```bash
pnpm add @mysten/sui @mysten/seal @mysten/walrus ai zod
```

--------------------------------

### Install manual client flow packages with pnpm

Source: https://github.com/mystenlabs/memwal/blob/dev/docs/getting-started/quick-start.md

Install packages required for the manual client flow using pnpm.

```bash
pnpm add @mysten/sui @mysten/seal @mysten/walrus
```

--------------------------------

### Clone Repo and Install Dependencies

Source: https://github.com/mystenlabs/memwal/blob/dev/docs/contributing/run-repo-locally.md

Clone the MemWal repository and install all necessary project dependencies using pnpm.

```bash
git clone https://github.com/CommandOSSLabs/MemWal.git
cd MemWal
pnpm install
```

--------------------------------

### Run the Relayer

Source: https://github.com/mystenlabs/memwal/blob/dev/docs/relayer/self-hosting.md

Steps to copy environment file, install dependencies, and run the relayer using Cargo.

```bash
cp services/server/.env.example services/server/.env
cd services/server/scripts
npm ci
cd ..
cargo run
```

--------------------------------

### Install Vercel AI SDK with npm

Source: https://github.com/mystenlabs/memwal/blob/dev/docs/getting-started/quick-start.md

Install the optional Vercel AI SDK package for AI middleware using npm.

```bash
npm install ai
```

--------------------------------

### Install Vercel CLI

Source: https://github.com/mystenlabs/memwal/blob/dev/apps/chatbot/README.md

Installs the Vercel Command Line Interface globally. This is a prerequisite for linking your local instance with Vercel and GitHub accounts.

```bash
npm i -g vercel
```

--------------------------------

### Install memwal with OpenAI integration

Source: https://github.com/mystenlabs/memwal/blob/dev/docs/python-sdk/quick-start.md

Install the memwal SDK with optional OpenAI integration. This enables memwal to function as middleware for OpenAI SDK usage.

```bash
pip install memwal[openai]
```

--------------------------------

### Install Walrus Memory SDK

Source: https://github.com/mystenlabs/memwal/blob/dev/SKILL.md

Install the Walrus Memory SDK using pnpm. Optional packages for Vercel AI SDK and manual client integration are also listed.

```bash
# Install the SDK
pnpm add @mysten-incubation/memwal

# Optional: for Vercel AI SDK integration
pnpm add ai zod

# Optional: for manual client (client-side SEAL encryption)
pnpm add @mysten/sui @mysten/seal @mysten/walrus
```

--------------------------------

### Start Server with Debug Logging and Namespace

Source: https://github.com/mystenlabs/memwal/blob/dev/packages/mcp/README.md

Manually start the Walrus Memory MCP server with debug logging enabled and a specific namespace. Useful for testing namespace injection.

```sh
MEMWAL_MCP_DEBUG=1 npx -y @mysten-incubation/memwal-mcp --namespace demo-ns
```

--------------------------------

### Install memwal with LangChain integration

Source: https://github.com/mystenlabs/memwal/blob/dev/docs/python-sdk/quick-start.md

Install the memwal SDK with optional LangChain integration. This allows using memwal as middleware within LangChain applications.

```bash
pip install memwal[langchain]
```

--------------------------------

### Example Benchmark Result Files

Source: https://github.com/mystenlabs/memwal/blob/dev/services/server/benchmarks/README.md

Provides a concrete example of the JSON artifact filenames produced after a specific LME (longmemeval) run. These filenames help identify specific benchmark runs and configurations.

```text
mem57-lme-20260520T144014Z-longmemeval-baseline.json
mem57-lme-20260520T144014Z-longmemeval-importance_heavy.json
mem57-lme-20260520T144014Z-longmemeval-session_map.json
```

--------------------------------

### Install MemWal SDK

Source: https://github.com/mystenlabs/memwal/blob/dev/docs/llms-full.txt

Install the MemWal SDK using pnpm. Optional peer dependencies for Vercel AI SDK middleware or manual client-side encryption are also listed.

```bash
pnpm add @mysten-incubation/memwal
```

```bash
# For Vercel AI SDK middleware
pnpm add ai zod

# For manual client (client-side SEAL encryption)
pnpm add @mysten/sui @mysten/seal @mysten/walrus
```

--------------------------------

### Install and Use Memwal LangChain Middleware

Source: https://github.com/mystenlabs/memwal/blob/dev/packages/python-sdk-memwal/notebooks/walrus_memory_python_sdk.ipynb

Installs the necessary packages and demonstrates how to wrap a LangChain chat model with Memwal's `with_memwal_langchain` function. This allows for memory management in LangChain applications. Ensure environment variables like OPENAI_API_KEY, MEMWAL_PRIVATE_KEY, MEMWAL_ACCOUNT_ID, WALRUS_MEMORY_NAMESPACE, WALRUS_MEMORY_ENV, OPENAI_MODEL, and OPENAI_BASE_URL are set.

```python
if RUN_MIDDLEWARE_DEMO:
    import subprocess
    import sys

    subprocess.check_call([
        sys.executable,
        "-m",
        "pip",
        "install",
        "-q",
        "-U",
        "memwal[langchain]",
        "langchain-openai",
    ])

    from langchain_core.messages import HumanMessage
    from langchain_openai import ChatOpenAI
    from memwal import with_memwal_langchain

    read_secret("OPENAI_API_KEY", "Paste OpenAI API key for LangChain demo (hidden): ")

    llm_kwargs = {"model": OPENAI_MODEL, "api_key": os.environ["OPENAI_API_KEY"]}
    if OPENAI_BASE_URL:
        llm_kwargs["base_url"] = OPENAI_BASE_URL
    llm = ChatOpenAI(**llm_kwargs)
    smart_llm = with_memwal_langchain(
        llm,
        key=os.environ["MEMWAL_PRIVATE_KEY"],
        account_id=os.environ["MEMWAL_ACCOUNT_ID"],
        namespace=WALRUS_MEMORY_NAMESPACE,
        env=WALRUS_MEMORY_ENV,
        max_memories=5,
        min_relevance=0.2,
    )

    response = await smart_llm.ainvoke([
        HumanMessage(content="Summarize the customer's support context in one sentence.")
    ])
    print(response.content)
else:
    print("LangChain middleware demo skipped. Set RUN_MIDDLEWARE_DEMO=true to run it.")

```

--------------------------------

### Build the Project

Source: https://github.com/mystenlabs/memwal/blob/dev/apps/noter/README.md

Run this command to build the project for deployment.

```bash
pnpm build
```

--------------------------------

### Verify MemWal Connection and Tools

Source: https://github.com/mystenlabs/memwal/blob/dev/docs/mcp/claude-code.md

Verify that MemWal is connected and its tools are available in Claude Code. This check confirms the installation and setup.

```bash
/mcp          → "memwal" should be Connected
```

--------------------------------

### Initialize and Use MemWal SDK

Source: https://github.com/mystenlabs/memwal/blob/dev/docs/fundamentals/architecture/core-components.md

Demonstrates how to create an instance of the MemWal SDK with configuration and perform a 'remember' operation. Ensure environment variables for private key, account ID, and server URL are set.

```typescript
const memwal = MemWal.create({
  key: process.env.MEMWAL_PRIVATE_KEY!,
  accountId: process.env.MEMWAL_ACCOUNT_ID!,
  serverUrl: process.env.MEMWAL_SERVER_URL,
  namespace: "my-app",
});

const job = await memwal.remember("User prefers dark mode.");
await memwal.waitForRememberJob(job.job_id);
```

--------------------------------

### Memwal SDK Quick Start

Source: https://github.com/mystenlabs/memwal/blob/dev/packages/sdk/README.md

Initialize the Memwal client and perform memory operations like remembering and recalling. Ensure environment variables for key, account ID, and server URL are set.

```typescript
import { MemWal } from "@mysten-incubation/memwal";

const memwal = MemWal.create({
  key: process.env.MEMWAL_PRIVATE_KEY!,
  accountId: process.env.MEMWAL_ACCOUNT_ID!,
  serverUrl: process.env.MEMWAL_SERVER_URL ?? "https://relayer.memory.walrus.xyz",
  namespace: "demo",
});

await memwal.rememberAndWait(
  "User prefers dark mode and uses TypeScript.",
  undefined,
  { timeoutMs: 30_000 },
);
const memories = await memwal.recall({
  query: "What are the user's preferences?",
  topK: 10,
  maxDistance: 0.7,
});
await memwal.restore("demo");

```

--------------------------------

### Setup and Use OpenAI Middleware with Memwal

Source: https://github.com/mystenlabs/memwal/blob/dev/packages/python-sdk-memwal/notebooks/walrus_memory_python_sdk.ipynb

Installs optional dependencies and configures an OpenAI client wrapped with Memwal middleware. This middleware injects relevant Walrus Memory context into chat prompts and can save the latest user message. Requires API keys and specific environment variables to be set.

```python
if RUN_MIDDLEWARE_DEMO:
    import subprocess
    import sys

    subprocess.check_call([sys.executable, "-m", "pip", "install", "-q", "-U", "memwal[openai]", "openai"])

    from openai import AsyncOpenAI
    from memwal import with_memwal_openai

    read_secret("OPENAI_API_KEY", "Paste OpenAI API key for middleware demo (hidden): ")

    openai_kwargs = {"api_key": os.environ["OPENAI_API_KEY"]}
    if OPENAI_BASE_URL:
        openai_kwargs["base_url"] = OPENAI_BASE_URL
    openai_client = AsyncOpenAI(**openai_kwargs)
    smart_openai = with_memwal_openai(
        openai_client,
        key=os.environ["MEMWAL_PRIVATE_KEY"],
        account_id=os.environ["MEMWAL_ACCOUNT_ID"],
        namespace=WALRUS_MEMORY_NAMESPACE,
        env=WALRUS_MEMORY_ENV,
        max_memories=5,
        min_relevance=0.2,
    )

    response = await smart_openai.chat.completions.create(
        model=OPENAI_MODEL,
        messages=[
            {
                "role": "user",
                "content": "Use my support memory: what TV issue and appointment window should the agent remember?",
            }
        ],
        max_tokens=160,
    )
    print(response.choices[0].message.content)
else:
    print("OpenAI middleware demo skipped. Set RUN_MIDDLEWARE_DEMO=true to run it.")

```

--------------------------------

### Async MemWal Client Example

Source: https://github.com/mystenlabs/memwal/blob/dev/docs/python-sdk/usage.md

Demonstrates the basic usage of the asynchronous MemWal client, including creating a client instance, remembering data, recalling data with parameters, and closing the client. Use this for async-native applications.

```python
from memwal import MemWal, RecallParams

memwal = MemWal.create(key="...", account_id="0x...", env="prod")
done = await memwal.remember_and_wait("User prefers dark mode.")
result = await memwal.recall(RecallParams(query="preferences"))
await memwal.close()
```

--------------------------------

### Install Codex hooks for MemWal plugin

Source: https://github.com/mystenlabs/memwal/blob/dev/packages/mcp/TESTING.md

Steps to install MemWal plugin hooks for Codex. This involves degitting the plugin and running an installation script. Ensure the installation script is run from a persistent directory.

```bash
npx degit MystenLabs/MemWal/packages/mcp/plugin /tmp/memwal-plugin
node /tmp/memwal-plugin/scripts/install_codex_hooks.mjs
```

--------------------------------

### MemWalManual Client Setup

Source: https://github.com/mystenlabs/memwal/blob/dev/docs/sdk/usage/memwal-manual.md

Initialize the MemWalManual client with necessary configuration including private keys, server URLs, API keys, and package details.

```typescript
import { MemWalManual } from "@mysten-incubation/memwal/manual";

const manual = MemWalManual.create({
  key: "<your-ed25519-private-key>",
  serverUrl: "https://your-relayer-url.com",
  suiPrivateKey: "<your-sui-private-key>",    // OR walletSigner
  embeddingApiKey: "<your-openai-api-key>",
  packageId: "<memwal-package-id>",
  accountId: "<memwal-account-id>",
  namespace: "chatbot-prod",
});
```

--------------------------------

### Install Walrus Memory Python SDK

Source: https://github.com/mystenlabs/memwal/blob/dev/packages/python-sdk-memwal/README.md

Install the memwal package using pip. Optional integrations for LangChain, OpenAI, or all packages can be installed using bracket notation.

```bash
pip install memwal
```

```bash
pip install memwal[langchain]   # LangChain support
pip install memwal[openai]      # OpenAI SDK support
pip install memwal[all]         # Everything
```

--------------------------------

### Install Claude Code plugin from marketplace

Source: https://github.com/mystenlabs/memwal/blob/dev/packages/mcp/TESTING.md

Commands to add and install the MemWal plugin for Claude Code. After installation, restart Claude Code and potentially log in against the dev relayer for testing.

```bash
/plugin marketplace add MystenLabs/MemWal
/plugin install memwal@memwal-plugins
```

--------------------------------

### Download Benchmark Dataset

Source: https://github.com/mystenlabs/memwal/blob/dev/services/server/benchmarks/README.md

Downloads the specified dataset (e.g., 'locomo') for use in the benchmarks. This is a one-time setup step per benchmark dataset.

```bash
python run.py download locomo
```

--------------------------------

### Install MemWal Python Package

Source: https://github.com/mystenlabs/memwal/blob/dev/docs/sdk/overview.md

Install the memwal Python package and optional middleware for LangChain and OpenAI.

```bash
pip install memwal
pip install memwal[langchain]   # LangChain middleware
pip install memwal[openai]      # OpenAI SDK middleware
pip install memwal[all]         # everything
```

--------------------------------

### Run Local Docs Development Server

Source: https://github.com/mystenlabs/memwal/blob/dev/docs/contributing/docs-workflow.md

Starts a local development server to preview documentation changes. Ensure this is run before making significant documentation updates.

```shell
pnpm dev:docs
```

--------------------------------

### Build Documentation Site

Source: https://github.com/mystenlabs/memwal/blob/dev/docs/contributing/docs-workflow.md

Builds the static documentation site. This command should be run to ensure the documentation is correctly generated before shipping.

```shell
pnpm build:docs
```

--------------------------------

### Install OpenClaw MemWal Plugin

Source: https://github.com/mystenlabs/memwal/blob/dev/README.md

Install the OpenClaw memory plugin for Walrus Memory using the openclaw CLI.

```bash
openclaw plugins install @mysten-incubation/oc-memwal
```

--------------------------------

### Install MemWal Plugin in Claude Code

Source: https://github.com/mystenlabs/memwal/blob/dev/docs/mcp/claude-code.md

Install the MemWal plugin after adding its marketplace. This enables automatic memory features.

```bash
/plugin install memwal@memwal-plugins
```

--------------------------------

### Using MemWal as an Async Context Manager

Source: https://github.com/mystenlabs/memwal/blob/dev/packages/python-sdk-memwal/README.md

Demonstrates how to use the MemWal SDK with an asynchronous context manager. This ensures that the client is properly initialized and closed, simplifying resource management.

```python
import os
from memwal import MemWal

async with MemWal.create(
    key=os.environ["MEMWAL_PRIVATE_KEY"],
    account_id=os.environ["MEMWAL_ACCOUNT_ID"],
) as memwal:
    await memwal.remember_and_wait("I prefer dark mode")
```

--------------------------------

### Async Quick Start with MemWal

Source: https://github.com/mystenlabs/memwal/blob/dev/packages/python-sdk-memwal/README.md

Demonstrates asynchronous operations for storing memories, recalling information, and analyzing conversations using the MemWal SDK. It includes waiting for background jobs to complete and closing the client connection.

```python
import asyncio
import os
from memwal import MemWal, RecallParams

async def main():
    memwal = MemWal.create(
        key=os.environ["MEMWAL_PRIVATE_KEY"],
        account_id=os.environ["MEMWAL_ACCOUNT_ID"],
        server_url=os.environ.get("MEMWAL_SERVER_URL", "https://relayer.memory.walrus.xyz"),
    )

    # Store a memory and wait until the background job is searchable
    result = await memwal.remember_and_wait("I'm allergic to peanuts")
    print(result.blob_id)

    # Recall memories
    matches = await memwal.recall(RecallParams(query="food allergies", limit=10, max_distance=0.7))
    for memory in matches.results:
        print(f"{memory.text} (relevance: {1 - memory.distance:.2f})")

    # Analyze conversation for facts and wait until extracted facts are searchable
    analysis = await memwal.analyze_and_wait("I love coffee and live in Tokyo")
    for fact in analysis.facts:
        print(fact.text)

    await memwal.close()

asyncio.run(main())
```

--------------------------------

### Initialize MemWal SDK

Source: https://github.com/mystenlabs/memwal/blob/dev/SKILL.md

Initialize the MemWal SDK with private key, account ID, server URL, and a namespace. Ensure environment variables are set for credentials.

```typescript
import { MemWal } from "@mysten-incubation/memwal";

const memwal = MemWal.create({
  key: process.env.MEMWAL_PRIVATE_KEY!,
  accountId: process.env.MEMWAL_ACCOUNT_ID!,
  serverUrl: process.env.MEMWAL_SERVER_URL ?? "https://relayer.memory.walrus.xyz",
  namespace: "my-app",
});
```

--------------------------------

### Run Development Servers

Source: https://github.com/mystenlabs/memwal/blob/dev/docs/contributing/run-repo-locally.md

Start development servers for different parts of the monorepo. Choose the command based on the component you need to run.

```bash
# Docs site (Mintlify)
pnpm dev:docs

# Demo apps (pick one)
pnpm dev:app          # Playground dashboard
pnpm dev:noter        # Note-taking app
pnpm dev:chatbot      # AI chatbot
pnpm dev:researcher   # Research assistant

# SDK in watch mode (recompiles on changes)
pnpm dev:sdk
```

--------------------------------

### First Memory with MemWal SDK

Source: https://github.com/mystenlabs/memwal/blob/dev/docs/llms-full.txt

Demonstrates how to initialize the MemWal SDK and store/retrieve a memory. Ensure you have a delegate key, account ID, and relayer URL.

```typescript
import { MemWal } from "@mysten-incubation/memwal";

const memwal = MemWal.create({
  key: "<your-ed25519-private-key>",
  accountId: "<your-memwal-account-id>",
  serverUrl: "https://relayer.memory.walrus.xyz",
  namespace: "demo",
});

await memwal.health();
const job = await memwal.remember("I live in Hanoi and prefer dark mode.");
await memwal.waitForRememberJob(job.job_id);

const result = await memwal.recall({ query: "What do we know about this user?" });
console.log(result.results);
```

--------------------------------

### Start Local Postgres Database

Source: https://github.com/mystenlabs/memwal/blob/dev/apps/status/README.md

Use this command to start the local Postgres history database using Docker Compose.

```sh
docker compose -f apps/status/docker-compose.yml up -d postgres
```

--------------------------------

### Initialize MemWal Client

Source: https://github.com/mystenlabs/memwal/blob/dev/docs/sdk/usage/memwal.md

Create an instance of the MemWal client with your private key, account ID, server URL, and namespace. Ensure you replace placeholder values with your actual credentials.

```typescript
import { MemWal } from "@mysten-incubation/memwal";

const memwal = MemWal.create({
  key: "<your-ed25519-private-key>",
  accountId: "<your-memwal-account-id>",
  serverUrl: "https://your-relayer-url.com",
  namespace: "chatbot-prod",
});
```

--------------------------------

### Committee Key Server Configuration Example

Source: https://github.com/mystenlabs/memwal/blob/dev/docs/relayer/self-hosting.md

Example of how to configure committee key servers using SEAL_SERVER_CONFIGS. This requires an aggregatorUrl.

```env
SEAL_SERVER_CONFIGS=[{"objectId":"0x...","weight":1,"aggregatorUrl":"https://seal-aggregator.example.com"}]
```

--------------------------------

### Install withMemWal Dependencies with yarn

Source: https://github.com/mystenlabs/memwal/blob/dev/docs/sdk/quick-start.md

Install dependencies for the withMemWal client using yarn. This client is for use with the Vercel AI SDK.

```bash
yarn add ai zod
```

--------------------------------

### Install withMemWal Dependencies with pnpm

Source: https://github.com/mystenlabs/memwal/blob/dev/docs/sdk/quick-start.md

Install dependencies for the withMemWal client using pnpm. This client is for use with the Vercel AI SDK.

```bash
pnpm add ai zod
```

--------------------------------

### Initialize MemWal Client and Perform Memory Operations

Source: https://github.com/mystenlabs/memwal/blob/dev/docs/examples/example-apps.md

Demonstrates creating a MemWal client and performing remember, recall, and analyze operations. This is useful for a full getting-started flow.

```typescript
const memwal = MemWal.create({
  key: delegateKey,
  accountId: accountObjectId,
  serverUrl,
  namespace,
});

const job = await memwal.remember(rememberText);
await memwal.waitForRememberJob(job.job_id);
await memwal.recall({ query: recallQuery, limit: 5 });
await memwal.analyze(analyzeText);
```

--------------------------------

### Install withMemWal Dependencies with npm

Source: https://github.com/mystenlabs/memwal/blob/dev/docs/sdk/quick-start.md

Install dependencies for the withMemWal client using npm. This client is for use with the Vercel AI SDK.

```bash
npm install ai zod
```

--------------------------------

### Copy Nautilus Configuration Templates

Source: https://github.com/mystenlabs/memwal/blob/dev/docs/relayer/nautilus-tee.md

Copy the example configuration files for Nautilus and runtime environment before making edits. These files serve as templates for your service's configuration.

```bash
cp services/server/deploy/nautilus/nautilus.toml.example .nautilus.toml
cp services/server/deploy/nautilus/runtime.env.example .env.nautilus
```

--------------------------------

### Install Vercel AI SDK with yarn

Source: https://github.com/mystenlabs/memwal/blob/dev/docs/getting-started/quick-start.md

Install the optional Vercel AI SDK package for AI middleware using yarn.

```bash
yarn add ai
```

--------------------------------

### Build the SDK

Source: https://github.com/mystenlabs/memwal/blob/dev/README.md

Compile the SDK, which is a prerequisite for other applications in the repository.

```bash
pnpm build:sdk
```

--------------------------------

### Create MemWal Client Instance

Source: https://github.com/mystenlabs/memwal/blob/dev/docs/python-sdk/usage/memwal.md

Instantiate the MemWal client for asynchronous operations. Configure with your private key, account ID, environment (or server URL), and a default namespace. For synchronous operations, use `MemWalSync.create(...)` with the same arguments.

```python
from memwal import MemWal, RecallParams

memwal = MemWal.create(
    key="<your-ed25519-private-key>",
    account_id="<your-memwal-account-id>",
    env="prod",            # or server_url="https://your-relayer-url.com"
    namespace="chatbot-prod",
)
```