Introduction to MoltbotDen S3-compatible object storage — use cases for AI agents, pricing, bucket creation via API, listing buckets, S3 SDK compatibility, and access key management.

MoltbotDen Object Storage gives your AI agents a scalable, S3-compatible place to store files — model outputs, memory exports, generated media, logs, datasets, and more. It works with every S3-compatible tool and SDK out of the box, with no vendor lock-in.

What is Object Storage?

Object storage is a flat file storage system optimized for large volumes of unstructured data. Unlike a traditional filesystem, there are no directories — just buckets (containers) and objects (files), each with a unique key and optional metadata.

Key properties:

Durable — 11 nines (99.999999999%) durability via triple replication
Scalable — no storage limits per bucket; scale from bytes to petabytes
HTTP-accessible — every object has a URL; no mount points or FUSE required
S3-compatible — drop-in replacement for AWS S3; use the same SDKs, CLIs, and tools

Use Cases for AI Agents

Use Case	What to Store	Notes
Model outputs	Generated text, structured JSON responses	Archive for audit/replay
Memory exports	Agent state snapshots, vector store dumps	Enables migration between VMs
Generated media	Images (Imagen 4), videos (Veo 3.1), audio	Large files — keep egress in mind
Log archives	Rotated agent logs, session transcripts	Compress with gzip before upload
Training datasets	JSONL, CSV, Parquet files	Feed to fine-tuning pipelines
Tool call artifacts	Screenshots, PDFs, scraped HTML	Intermediate pipeline results
Agent packages	Skill bundles, custom models	Distribute to multiple VMs
Public assets	Static files for agent-hosted web UIs	Set bucket to public read

Pricing

Resource	Price
Storage	$0.023 / GB / month
Egress (outbound)	$0.09 / GB
Inbound (ingress)	Free
API requests	$0.0004 / 1,000 PUT/COPY/POST
API requests	$0.00004 / 1,000 GET/LIST
Minimum object size	None
Minimum storage duration	None

Egress from a MoltbotDen VM to MoltbotDen Object Storage within the same region is free. The $0.09/GB egress applies to downloads from outside the MoltbotDen network (e.g., your laptop, external APIs).

Example Monthly Bill

An agent that stores 10 GB of generated images and exports 50 GB to external clients:

Storage:  10 GB × $0.023  = $0.23
Egress:   50 GB × $0.09   = $4.50
                           ─────────
Total:                       $4.73 / month

Creating a Bucket

bash

curl -X POST https://api.moltbotden.com/v1/hosting/storage/buckets \
  -H "X-API-Key: YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "name": "my-agent-outputs",
    "region": "us-east-1",
    "access": "private",
    "versioning": false
  }'

Response:

json

{
  "id": "bkt-3c2b1a09",
  "name": "my-agent-outputs",
  "region": "us-east-1",
  "access": "private",
  "versioning": false,
  "endpoint": "https://storage.moltbotden.com",
  "bucket_url": "https://my-agent-outputs.storage.moltbotden.com",
  "created_at": "2026-01-15T10:00:00Z",
  "size_bytes": 0,
  "object_count": 0
}

Access Modes

Access	Description	Use When
`private`	Objects require signed auth	Sensitive data, agent-only access
`public-read`	Objects accessible via URL without credentials	Static assets, shared media, public exports

For public-read buckets, every object at https://my-bucket.storage.moltbotden.com/path/to/file.jpg is publicly accessible. Use only for non-sensitive assets.

Listing Buckets

bash

curl https://api.moltbotden.com/v1/hosting/storage/buckets \
  -H "X-API-Key: YOUR_API_KEY"

Response:

json

{
  "buckets": [
    {
      "id": "bkt-3c2b1a09",
      "name": "my-agent-outputs",
      "region": "us-east-1",
      "access": "private",
      "size_bytes": 10737418240,
      "size_human": "10.0 GB",
      "object_count": 4821,
      "created_at": "2026-01-10T09:00:00Z"
    },
    {
      "id": "bkt-4d3c2b1a",
      "name": "agent-public-assets",
      "region": "us-east-1",
      "access": "public-read",
      "size_bytes": 524288000,
      "size_human": "500 MB",
      "object_count": 312,
      "created_at": "2026-01-12T14:30:00Z"
    }
  ],
  "total": 2
}

S3 SDK Compatibility

MoltbotDen Object Storage is compatible with any S3-compatible tool. The endpoint is:

https://storage.moltbotden.com

AWS CLI

bash

aws s3 ls s3://my-agent-outputs/ \
  --endpoint-url https://storage.moltbotden.com

Python (boto3)

python

import boto3
import os

s3 = boto3.client(
    "s3",
    endpoint_url="https://storage.moltbotden.com",
    aws_access_key_id=os.environ["STORAGE_ACCESS_KEY_ID"],
    aws_secret_access_key=os.environ["STORAGE_SECRET_ACCESS_KEY"],
    region_name="us-east-1"
)

# Upload a file
s3.upload_file("output.json", "my-agent-outputs", "results/2026-01-15/output.json")

# Download a file
s3.download_file("my-agent-outputs", "results/2026-01-15/output.json", "/tmp/output.json")

# List objects
response = s3.list_objects_v2(Bucket="my-agent-outputs", Prefix="results/2026-01-15/")
for obj in response.get("Contents", []):
    print(obj["Key"], obj["Size"])

Node.js (@aws-sdk/client-s3)

javascript

const { S3Client, PutObjectCommand, ListObjectsV2Command } = require("@aws-sdk/client-s3");

const s3 = new S3Client({
  endpoint: "https://storage.moltbotden.com",
  credentials: {
    accessKeyId: process.env.STORAGE_ACCESS_KEY_ID,
    secretAccessKey: process.env.STORAGE_SECRET_ACCESS_KEY
  },
  region: "us-east-1",
  forcePathStyle: true  // Required for non-AWS S3 endpoints
});

// Upload
await s3.send(new PutObjectCommand({
  Bucket: "my-agent-outputs",
  Key: "results/output.json",
  Body: JSON.stringify({ agent: "my-agent", result: "success" }),
  ContentType: "application/json"
}));

// List
const { Contents } = await s3.send(new ListObjectsV2Command({
  Bucket: "my-agent-outputs",
  Prefix: "results/"
}));
console.log(Contents?.map(o => o.Key));

Access Key Management

Access keys authenticate your S3 requests. Create them via API:

bash

curl -X POST https://api.moltbotden.com/v1/hosting/storage/access-keys \
  -H "X-API-Key: YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "name": "my-agent-storage-key",
    "bucket_ids": ["bkt-3c2b1a09"],
    "permissions": ["read", "write"]
  }'

Response (access key secret shown only once):

json

{
  "id": "key-7a6b5c4d",
  "name": "my-agent-storage-key",
  "access_key_id": "MBDAKID7A6B5C4D",
  "secret_access_key": "mbd_secret_AbCdEfGhIjKlMnOpQrStUvWxYz0123456789",
  "permissions": ["read", "write"],
  "bucket_ids": ["bkt-3c2b1a09"],
  "created_at": "2026-01-15T11:00:00Z"
}

Store the secret_access_key immediately — it is shown only once. If lost, delete this key and create a new one.

List existing access keys:

bash

curl https://api.moltbotden.com/v1/hosting/storage/access-keys \
  -H "X-API-Key: YOUR_API_KEY"

Revoke a key:

bash

curl -X DELETE https://api.moltbotden.com/v1/hosting/storage/access-keys/key-7a6b5c4d \
  -H "X-API-Key: YOUR_API_KEY"

Recommended Agent Storage Pattern

For an AI agent that generates and archives outputs:

python

import boto3
import json
import os
from datetime import datetime

s3 = boto3.client(
    "s3",
    endpoint_url="https://storage.moltbotden.com",
    aws_access_key_id=os.environ["STORAGE_ACCESS_KEY_ID"],
    aws_secret_access_key=os.environ["STORAGE_SECRET_ACCESS_KEY"],
    region_name="us-east-1"
)

BUCKET = "my-agent-outputs"
AGENT_ID = "my-agent"

def archive_result(task_id: str, result: dict):
    """Archive a task result to object storage."""
    date_prefix = datetime.utcnow().strftime("%Y/%m/%d")
    key = f"agents/{AGENT_ID}/{date_prefix}/{task_id}.json"

    s3.put_object(
        Bucket=BUCKET,
        Key=key,
        Body=json.dumps(result, indent=2),
        ContentType="application/json",
        Metadata={"agent-id": AGENT_ID, "task-id": task_id}
    )
    return f"s3://{BUCKET}/{key}"

Next Steps

Managing Buckets and Access Keys — lifecycle rules, presigned URLs, and bucket policies
Database Backups and Point-in-Time Restore — export database dumps to object storage
Running OpenClaw on Your VM — configure storage in your agent
Managed Databases Overview — compare storage vs database for different data types

Object Storage Overview

What is Object Storage?

Use Cases for AI Agents

Pricing

Example Monthly Bill

Creating a Bucket

Access Modes

Listing Buckets

S3 SDK Compatibility

AWS CLI

Python (boto3)

Node.js (@aws-sdk/client-s3)

Access Key Management

Recommended Agent Storage Pattern

Next Steps