### MCP Integration - Endpoint Setup

Source: https://context7.com/localrivet/datasaver/llms.txt

Instructions on how to enable and access the MCP endpoint for programmatic backup management using Bearer token authentication.

```APIDOC
## MCP Integration - Endpoint Setup

The MCP (Model Context Protocol) endpoint allows AI assistants like Claude to manage backups programmatically with Bearer token authentication.

### Setup
```bash
# Enable MCP endpoint
export DATASAVER_MCP_API_KEY="your-secure-api-key"
export DATASAVER_BASE_URL="https://your-domain.com"
datasaver daemon
```

### MCP Endpoint URL
`http://localhost:8080/mcp`

### OAuth Discovery Endpoints
- `GET /.well-known/oauth-protected-resource`
- `GET /.well-known/oauth-authorization-server`
```

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

### YAML Configuration Example for DataSaver

Source: https://github.com/localrivet/datasaver/blob/main/docs/configuration.md

This snippet shows an example of a `config.yaml` file used to configure DataSaver. It includes settings for database, storage, schedule, retention, backup verification, and monitoring. This file needs to be created and referenced when running the `datasaver daemon` command.

```yaml
database:
  type: postgres
  host: localhost
  port: 5432
  name: myapp
  user: backup_user
  password: secret

storage:
  backend: s3
  path: ./backups
  s3:
    bucket: my-backups
    endpoint: s3.amazonaws.com
    region: us-west-2
    access_key: AKIAIOSFODNN7EXAMPLE
    secret_key: wJalrXUtnFEMI/K7MDENG/bPxRfiCYEXAMPLEKEY

schedule: "0 */6 * * *"  # Every 6 hours

retention:
  daily: 7
  weekly: 4
  monthly: 12

backup:
  verify_after_backup: true
  verify_checksum: true

monitoring:
  health_port: 8080
  metrics_port: 9090
  webhook_url: https://hooks.slack.com/services/...
  alert_after_hours: 26

```

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

### MCP Endpoint Setup (Bash)

Source: https://context7.com/localrivet/datasaver/llms.txt

Shell commands to enable and configure the MCP (Model Context Protocol) endpoint for programmatic backup management. It requires setting environment variables for API key and base URL, then running the datasaver daemon.

```bash
# Enable MCP endpoint
export DATASAVER_MCP_API_KEY="your-secure-api-key"
export DATASAVER_BASE_URL="https://your-domain.com"
datasaver daemon

# MCP endpoint available at: http://localhost:8080/mcp

# OAuth discovery endpoints:
# GET /.well-known/oauth-protected-resource
# GET /.well-known/oauth-authorization-server
```

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

### Go Backup Engine API Usage (Go)

Source: https://context7.com/localrivet/datasaver/llms.txt

Example Go program demonstrating the usage of the datasaver backup engine API. It covers loading configuration, setting up storage and notifiers, running backups, listing existing backups, and cleaning up old ones.

```go
package main

import (
    "context"
    "log/slog"
    "os"

    "github.com/localrivet/datasaver/internal/backup"
    "github.com/localrivet/datasaver/internal/config"
    "github.com/localrivet/datasaver/internal/notify"
    "github.com/localrivet/datasaver/internal/storage"
)

func main() {
    ctx := context.Background()
    logger := slog.New(slog.NewJSONHandler(os.Stdout, nil))

    // Load configuration
    cfg, err := config.Load("config.yaml")
    if err != nil {
        logger.Error("failed to load config", "error", err)
        os.Exit(1)
    }

    // Create storage backend
    factory := storage.NewFactory()
    store, err := factory.Create(cfg.Storage.Backend, cfg.Storage.Path, nil)
    if err != nil {
        logger.Error("failed to create storage", "error", err)
        os.Exit(1)
    }

    // Create notifier (optional)
    notifier := notify.NewNotifier(cfg.Monitoring.WebhookURL, logger)

    // Create backup engine
    engine := backup.NewEngine(cfg, store, notifier, logger)

    // Run backup
    result, err := engine.Run(ctx)
    if err != nil {
        logger.Error("backup failed", "error", err)
        os.Exit(1)
    }

    logger.Info("backup completed",
        "id", result.ID,
        "size", result.Size,
        "compressed_size", result.CompressedSize,
        "duration", result.Duration,
    )

    // List backups
    backups, _ := engine.ListBackups(ctx)
    for _, b := range backups {
        logger.Info("backup", "id", b.ID, "timestamp", b.Timestamp, "type", b.Type)
    }

    // Cleanup old backups
    deleted, _ := engine.Cleanup(ctx)
    logger.Info("cleanup completed", "deleted", deleted)
}
```

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

### Scheduler API (Go)

Source: https://context7.com/localrivet/datasaver/llms.txt

Provides cron-based backup scheduling with manual trigger support. Allows starting the scheduler, checking the next run time, triggering immediate backups, and stopping gracefully. Requires configuration, storage, and logging.

```go
package main

import (
    "context"
    "log/slog"
    "os"
    "os/signal"
    "syscall"

    "github.com/localrivet/datasaver/internal/backup"
    "github.com/localrivet/datasaver/internal/config"
    "github.com/localrivet/datasaver/internal/storage"
)

func main() {
    ctx, cancel := context.WithCancel(context.Background())
    defer cancel()
    logger := slog.New(slog.NewJSONHandler(os.Stdout, nil))

    cfg, _ := config.Load("")
    factory := storage.NewFactory()
    store, _ := factory.Create("local", "/backups", nil)

    engine := backup.NewEngine(cfg, store, nil, logger)
    scheduler := backup.NewScheduler(engine, "0 2 * * *", logger)

    // Start scheduler
    scheduler.Start(ctx)
    // Logs: scheduler started, schedule: 0 2 * * *, next_run: 2024-01-16T02:00:00Z

    // Check next scheduled run
    nextRun := scheduler.NextRun()
    // nextRun: 2024-01-16T02:00:00Z

    // Trigger immediate backup
    result, _ := scheduler.RunNow(ctx)
    // result.ID: backup_20240115_143022

    // Check if running
    running := scheduler.IsRunning()
    // running: true

    // Wait for shutdown signal
    sigCh := make(chan os.Signal, 1)
    signal.Notify(sigCh, syscall.SIGINT, syscall.SIGTERM)
    <-sigCh

    // Stop scheduler gracefully
    scheduler.Stop()
}
```

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

### Database Driver Interface (Go)

Source: https://context7.com/localrivet/datasaver/llms.txt

Abstracts database operations for PostgreSQL and SQLite. Supports connecting, getting version, and dumping database content to a writer. Requires database connection configurations.

```go
package main

import (
    "context"
    "os"

    "github.com/localrivet/datasaver/pkg/database"
)

func main() {
    ctx := context.Background()

    // PostgreSQL driver
    pgDriver, _ := database.NewDriver(database.Config{
        Type:     "postgres",
        Host:     "localhost",
        Port:     5432,
        Name:     "myapp",
        User:     "postgres",
        Password: "secret",
    })
    defer pgDriver.Close()

    pgDriver.Connect(ctx)
    version, _ := pgDriver.Version(ctx)
    // version: "16.1"

    // Dump to writer
    f, _ := os.Create("backup.dump")
    pgDriver.Dump(ctx, f)
    f.Close()

    // SQLite driver
    sqliteDriver, _ := database.NewDriver(database.Config{
        Type: "sqlite",
        Path: "/data/myapp.db",
    })
    defer sqliteDriver.Close()

    sqliteDriver.Connect(ctx)
    sqliteDriver.Dump(ctx, os.Stdout)
}
```

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

### S3 Storage Backend Setup for Datasaver

Source: https://context7.com/localrivet/datasaver/llms.txt

Configures Datasaver to store backups in S3-compatible storage, such as AWS S3 or MinIO. This setup requires the Datasaver image and environment variables specifying S3 bucket details, endpoint, region, credentials, and SSL usage. Compression can also be configured.

```yaml
# compose.yml with S3 storage
services:
  datasaver:
    image: ghcr.io/localrivet/datasaver:latest
    environment:
      DATASAVER_DB_HOST: postgres
      DATASAVER_DB_NAME: myapp
      DATASAVER_DB_USER: postgres
      DATASAVER_DB_PASSWORD: secret
      DATASAVER_SCHEDULE: "0 */6 * * *"
      DATASAVER_STORAGE_BACKEND: s3
      DATASAVER_S3_BUCKET: my-backups
      DATASAVER_S3_ENDPOINT: s3.amazonaws.com
      DATASAVER_S3_REGION: us-west-2
      DATASAVER_S3_ACCESS_KEY: ${AWS_ACCESS_KEY_ID}
      DATASAVER_S3_SECRET_KEY: ${AWS_SECRET_ACCESS_KEY}
      DATASAVER_S3_USE_SSL: "true"
      DATASAVER_COMPRESSION: gzip
    depends_on:
      - postgres
```

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

### Go Library - Backup Engine API

Source: https://context7.com/localrivet/datasaver/llms.txt

Usage example of the Datasaver Go library's backup engine for programmatic backup operations.

```APIDOC
## Go Library - Backup Engine API

The backup engine provides programmatic access to backup operations for custom integrations.

### Example Usage
```go
package main

import (
    "context"
    "log/slog"
    "os"

    "github.com/localrivet/datasaver/internal/backup"
    "github.com/localrivet/datasaver/internal/config"
    "github.com/localrivet/datasaver/internal/notify"
    "github.com/localrivet/datasaver/internal/storage"
)

func main() {
    ctx := context.Background()
    logger := slog.New(slog.NewJSONHandler(os.Stdout, nil))

    // Load configuration
    cfg, err := config.Load("config.yaml")
    if err != nil {
        logger.Error("failed to load config", "error", err)
        os.Exit(1)
    }

    // Create storage backend
    factory := storage.NewFactory()
    store, err := factory.Create(cfg.Storage.Backend, cfg.Storage.Path, nil)
    if err != nil {
        logger.Error("failed to create storage", "error", err)
        os.Exit(1)
    }

    // Create notifier (optional)
    notifier := notify.NewNotifier(cfg.Monitoring.WebhookURL, logger)

    // Create backup engine
    engine := backup.NewEngine(cfg, store, notifier, logger)

    // Run backup
    result, err := engine.Run(ctx)
    if err != nil {
        logger.Error("backup failed", "error", err)
        os.Exit(1)
    }

    logger.Info("backup completed",
        "id", result.ID,
        "size", result.Size,
        "compressed_size", result.CompressedSize,
        "duration", result.Duration,
    )

    // List backups
    backups, _ := engine.ListBackups(ctx)
    for _, b := range backups {
        logger.Info("backup", "id", b.ID, "timestamp", b.Timestamp, "type", b.Type)
    }

    // Cleanup old backups
    deleted, _ := engine.Cleanup(ctx)
    logger.Info("cleanup completed", "deleted", deleted)
}
```
```

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

### MCP Integration - Tool Calls

Source: https://context7.com/localrivet/datasaver/llms.txt

Available MCP tool calls for managing backups, including triggering backups, listing, restoring, verifying, and getting system status.

```APIDOC
## MCP Integration - Tool Calls

Available MCP tools for backup management through AI assistants.

### `backup_now`
**Description**: Trigger an immediate backup.
**Arguments**: None

### `list_backups`
**Description**: List available backups.
**Arguments**:
- `limit` (integer) - Optional - The maximum number of backups to return.

### `restore_backup`
**Description**: Restore from a backup.
**Arguments**:
- `backup_id` (string) - Required - The ID of the backup to restore.
- `target_db` (string) - Required - The name of the target database for restoration.
- `dry_run` (boolean) - Optional - If true, perform a dry run without actually restoring.

### `verify_backup`
**Description**: Verify the integrity of a backup.
**Arguments**:
- `backup_id` (string) - Required - The ID of the backup to verify.

### `get_status`
**Description**: Get the system status.
**Arguments**: None
```

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

### Start datasaver Daemon (Bash)

Source: https://github.com/localrivet/datasaver/blob/main/docs/mcp-integration.md

Starts the datasaver daemon, which exposes the MCP endpoint. The endpoint will then be accessible at http://localhost:8080/mcp by default.

```bash
datasaver daemon
```

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

### Basic PostgreSQL Sidecar Setup with Local Storage

Source: https://context7.com/localrivet/datasaver/llms.txt

Deploys Datasaver as a sidecar container alongside a PostgreSQL database for automated backups using local storage. It requires the PostgreSQL image and Datasaver image, along with environment variables for database credentials and backup schedules. Volumes are defined for persistent PostgreSQL data and backup storage.

```yaml
# compose.yml
services:
  postgres:
    image: postgres:16
    environment:
      POSTGRES_DB: myapp
      POSTGRES_USER: postgres
      POSTGRES_PASSWORD: secret
    volumes:
      - postgres_data:/var/lib/postgresql/data

  datasaver:
    image: ghcr.io/localrivet/datasaver:latest
    environment:
      DATASAVER_DB_HOST: postgres
      DATASAVER_DB_NAME: myapp
      DATASAVER_DB_USER: postgres
      DATASAVER_DB_PASSWORD: secret
      DATASAVER_SCHEDULE: "0 2 * * *"
      DATASAVER_STORAGE_BACKEND: local
      DATASAVER_STORAGE_PATH: /backups
      DATASAVER_KEEP_DAILY: 7
      DATASAVER_KEEP_WEEKLY: 4
      DATASAVER_KEEP_MONTHLY: 6
    volumes:
      - backup_data:/backups
    depends_on:
      - postgres
    restart: unless-stopped

volumes:
  postgres_data:
  backup_data:
```

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

### MCP Tool Calls (JSON)

Source: https://context7.com/localrivet/datasaver/llms.txt

Examples of JSON payloads representing tool calls for managing backups via the MCP endpoint. These include actions like triggering backups, listing, restoring, verifying, and checking system status.

```json
// backup_now - Trigger immediate backup
{
  "name": "backup_now",
  "arguments": {}
}
```

```json
// list_backups - List available backups
{
  "name": "list_backups",
  "arguments": {
    "limit": 10
  }
}
```

```json
// restore_backup - Restore from backup
{
  "name": "restore_backup",
  "arguments": {
    "backup_id": "backup_20240115_0200",
    "target_db": "myapp_restored",
    "dry_run": true
  }
}
```

```json
// verify_backup - Verify backup integrity
{
  "name": "verify_backup",
  "arguments": {
    "backup_id": "backup_20240115_0200"
  }
}
```

```json
// get_status - Get system status
{
  "name": "get_status",
  "arguments": {}
}
```

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

### datasaver CLI Commands for Management and Restore

Source: https://github.com/localrivet/datasaver/blob/main/README.md

Provides examples of datasaver CLI commands for daemon operation, listing backups, performing restores (including to a different database and dry-run), cleanup, health checks, and verifying backup integrity. These commands offer granular control over backup management.

```bash
### `datasaver daemon`
# Run as scheduled backup daemon. Starts the scheduler, health endpoint, and metrics server.
```bash
datasaver daemon
datasaver daemon --config /etc/datasaver/config.yml
```

### `datasaver backup`
# Perform an immediate backup.
```bash
datasaver backup
```

### `datasaver list`
# List all available backups.
```bash
datasaver list
```

### `datasaver restore <backup-id>`
# Restore from a specific backup.
```bash
datasaver restore backup_20240111_0200

# Restore to different database
datasaver restore backup_20240111_0200 --target-db mydb_restored

# Dry run (test without applying)
datasaver restore backup_20240111_0200 --dry-run
```

### `datasaver cleanup`
# Manually run the cleanup routine to delete old backups.
```bash
datasaver cleanup
```

### `datasaver health`
# Check backup system health.
```bash
datasaver health
```

### `datasaver verify <backup-id>`
# Validate backup integrity.
```bash
datasaver verify backup_20240111_0200
```
```

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

### Configuring Datasaver Webhook Notifications

Source: https://context7.com/localrivet/datasaver/llms.txt

This example demonstrates how to configure Datasaver to send webhook notifications to a specified URL upon backup events, such as success, failure, or alerts. The DATASAVER_WEBHOOK_URL environment variable is used to set the target URL for these notifications.

```bash
# Configure webhook
export DATASAVER_WEBHOOK_URL=https://hooks.example.com/backup
```

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

### datasaver Configuration File (YAML)

Source: https://github.com/localrivet/datasaver/blob/main/README.md

Example YAML configuration file for datasaver. This allows for structured configuration of database connection, schedule, storage (including S3 details), retention policies, compression, and monitoring parameters.

```yaml
database:
  url: "postgres://user:pass@postgres:5432/mydb"

schedule: "0 2 * * *"

storage:
  backend: s3
  s3:
    bucket: my-backups
    endpoint: s3.amazonaws.com
    region: us-east-1
    access_key: ${AWS_ACCESS_KEY}
    secret_key: ${AWS_SECRET_KEY}

retention:
  daily: 7
  weekly: 4
  monthly: 6
  max_age_days: 90

compression: gzip

monitoring:
  metrics_port: 9090
  health_port: 8080
  webhook_url: https://hooks.example.com/backup
  alert_after_hours: 26
```

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

### Querying Datasaver Prometheus Metrics Endpoint

Source: https://context7.com/localrivet/datasaver/llms.txt

This command retrieves Prometheus metrics exposed by Datasaver, providing insights into backup performance and system usage. Key metrics include backup duration, size, success rate, and storage utilization. The example also includes a Prometheus scrape configuration.

```bash
# Query metrics endpoint
curl http://localhost:9090/metrics

# Key metrics exposed:
# datasaver_backup_duration_seconds - Histogram of backup duration
# datasaver_backup_size_bytes - Gauge of last backup size
# datasaver_backups_total - Counter of total backup attempts
# datasaver_backup_failures_total - Counter of failed backups
# datasaver_last_backup_timestamp - Gauge of last backup time
# datasaver_last_backup_success - Gauge (1=success, 0=failure)
# datasaver_storage_used_bytes - Gauge of total storage used

# Example Prometheus scrape config:
# scrape_configs:
#   - job_name: 'datasaver'
#     static_configs:
#       - targets: ['datasaver:9090']
```

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

### Docker Compose Configuration for datasaver

Source: https://github.com/localrivet/datasaver/blob/main/README.md

Example Docker Compose configuration to deploy datasaver as a sidecar container. It specifies environment variables for database connection, scheduling, storage, and retention policies. Ensure `backup_data` volume is correctly configured for persistent storage.

```yaml
datasaver:
  image: ghcr.io/localrivet/datasaver:latest
  environment:
    DATASAVER_DB_HOST: postgres
    DATASAVER_DB_NAME: mydb
    DATASAVER_DB_USER: postgres
    DATASAVER_DB_PASSWORD: secret
    DATASAVER_SCHEDULE: "0 2 * * *"
    DATASAVER_STORAGE_BACKEND: local
    DATASAVER_STORAGE_PATH: /backups
    DATASAVER_KEEP_DAILY: 7
    DATASAVER_KEEP_WEEKLY: 4
    DATASAVER_KEEP_MONTHLY: 6
  volumes:
    - backup_data:/backups
  depends_on:
    - postgres
  restart: unless-stopped
```

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

### Example Webhook Notification Payload (JSON)

Source: https://github.com/localrivet/datasaver/blob/main/README.md

This JSON object represents the payload sent to a configured webhook URL upon backup events. It includes event details such as the event type, timestamp, backup ID, status, and specific details about the backup's size and duration. This format is crucial for integrating datasaver events with other systems.

```json
{
  "event": "backup.completed",
  "timestamp": "2024-01-11T02:00:15Z",
  "backup_id": "backup_20240111_0200",
  "status": "success",
  "message": "Backup backup_20240111_0200 completed successfully",
  "details": {
    "size_bytes": 131621888,
    "duration_ms": 12500
  }
}
```

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

### Storage Backend Interface (Go)

Source: https://context7.com/localrivet/datasaver/llms.txt

Provides abstraction for local filesystem and S3-compatible storage. Supports writing, reading, listing, checking existence, and deleting backup files. Requires storage specific configurations.

```go
package main

import (
    "bytes"
    "context"
    "io"

    "github.com/localrivet/datasaver/internal/storage"
)

func main() {
    ctx := context.Background()

    // Local storage
    localStore, _ := storage.NewLocalStorage("/backups")

    // Write backup
    data := bytes.NewReader([]byte("backup data"))
    localStore.Write(ctx, "backup_20240115.dump", data)

    // Read backup
    reader, _ := localStore.Read(ctx, "backup_20240115.dump")
    content, _ := io.ReadAll(reader)
    reader.Close()

    // List backups
    files, _ := localStore.List(ctx, "")
    for _, f := range files {
        // f.Path, f.Size, f.LastModified
    }

    // Check existence
    exists, _ := localStore.Exists(ctx, "backup_20240115.dump")

    // Delete backup
    localStore.Delete(ctx, "backup_20240115.dump")

    // S3 storage
    s3Store, _ := storage.NewS3Storage(storage.S3Config{
        Bucket:    "my-backups",
        Endpoint:  "s3.amazonaws.com",
        Region:    "us-west-2",
        AccessKey: "AKIAIOSFODNN7EXAMPLE",
        SecretKey: "wJalrXUtnFEMI/K7MDENG/bPxRfiCYEXAMPLEKEY",
        UseSSL:    true,
    })
    // Same interface as local storage
    s3Store.Write(ctx, "backup_20240115.dump", bytes.NewReader(content))
}
```

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

### Datasaver Application Structure (Go)

Source: https://github.com/localrivet/datasaver/blob/main/prd.md

Illustrates the directory structure for the datasaver Go application, highlighting the organization of core logic, modules, and entry points. Key directories include 'cmd' for the CLI entry point, 'internal' for application logic, and 'pkg' for reusable packages.

```tree
datasaver/
├── cmd/
│   └── datasaver/
│       └── main.go           # CLI entry point
├── internal/
│   ├── backup/
│   │   ├── engine.go         # Core backup logic
│   │   ├── scheduler.go      # Cron scheduling
│   │   └── validator.go      # Backup validation
│   ├── storage/
│   │   ├── interface.go      # Storage abstraction
│   │   ├── local.go          # Local filesystem
│   │   └── s3.go             # S3-compatible storage
│   ├── rotation/
│   │   ├── policy.go         # Retention policies
│   │   └── gfs.go            # GFS rotation logic
│   ├── restore/
│   │   ├── engine.go         # Restore operations
│   │   └── pitr.go           # Point-in-time recovery
│   ├── config/
│   │   └── config.go         # Configuration management
│   ├── metrics/
│   │   └── prometheus.go     # Metrics collection
│   └── notify/
│       └── webhook.go        # Notification handling
├── pkg/
│   └── postgres/
│       ├── connection.go     # Postgres client
│       └── metadata.go       # Backup metadata
├── Dockerfile
├── docker-compose.yml        # Example deployment
└── README.md
```

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

### Build and Run datasaver Application (Bash)

Source: https://github.com/localrivet/datasaver/blob/main/CLAUDE.md

These commands demonstrate how to build the datasaver Go application binary, run its tests, create a Docker image, and launch it using docker-compose. This is essential for development and deployment.

```bash
# Build binary
go build -o datasaver ./cmd/datasaver

# Run tests
go test ./...

# Build Docker image
docker build -t datasaver .

# Run with docker-compose
docker compose up -d
```

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

### Build Binary and Docker Image (Bash)

Source: https://github.com/localrivet/datasaver/blob/main/README.md

These bash commands are used to build the datasaver project. The first command compiles the Go source code into an executable binary named 'datasaver'. The second command builds a Docker image tagged as 'datasaver' from the project's Dockerfile, enabling containerized deployment.

```bash
# Build binary
go build -o datasaver ./cmd/datasaver

# Build Docker image
docker build -t datasaver .
```

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

### Datasaver CLI - Daemon Mode (Bash)

Source: https://github.com/localrivet/datasaver/blob/main/prd.md

Demonstrates how to run the datasaver application in daemon mode using the CLI. The primary command is `datasaver daemon`. An optional flag `--config` can be used to specify a custom configuration file path, allowing for flexible deployment configurations.

```bash
# Run as scheduled backup daemon
datasaver daemon

# Run with config file
datasaver daemon --config /etc/datasaver/config.yml
```

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

### Datasaver CLI - One-Time Operations (Bash)

Source: https://github.com/localrivet/datasaver/blob/main/prd.md

Details various one-time operations that can be performed using the datasaver CLI. This includes initiating an immediate backup (`backup`), listing available backups with their details (`list`), restoring backups with options for target database and dry-run (`restore`), manually cleaning up old backups (`cleanup`), and verifying backup integrity (`verify`). The `list` command shows a sample output format.

```bash
# Perform immediate backup
datasaver backup

# List available backups
datasaver list
# Output:
# ID                    DATE                SIZE      TYPE
# backup_20240111_0200  2024-01-11 02:00   125.5 MB  daily
# backup_20240110_0200  2024-01-10 02:00   124.8 MB  daily

# Restore from backup
datasaver restore backup_20240111_0200

# Restore to different database
datasaver restore backup_20240111_0200 --target-db mydb_restored

# Test restore without applying
datasaver restore backup_20240111_0200 --dry-run

# Clean up old backups manually
datasaver cleanup

# Validate backup integrity
datasaver verify backup_20240111_0200
```

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

### Cleanup Algorithm for Backup Rotation

Source: https://github.com/localrivet/datasaver/blob/main/prd.md

Outlines the step-by-step process for cleaning up old backups based on defined retention policies. It involves scanning all backups, sorting by date, applying retention counts for daily, weekly, and monthly types, filtering by max age, and finally deleting marked backups.

```text
1. Scan all backups
2. For each backup type (daily/weekly/monthly):
   - Sort by date descending
   - Keep first N according to retention policy
   - Mark others for deletion if not protected by another policy
3. Apply max_age_days filter
4. Delete marked backups
```

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

### Claude Desktop Configuration

Source: https://context7.com/localrivet/datasaver/llms.txt

Configuration JSON for Claude Desktop to connect to the Datasaver MCP endpoint.

```APIDOC
## Claude Desktop Configuration

Configure Claude Desktop to connect to the datasaver MCP endpoint.

```json
// claude_desktop_config.json
{
  "mcpServers": {
    "datasaver": {
      "url": "http://localhost:8080/mcp",
      "headers": {
        "Authorization": "Bearer your-secure-api-key"
      }
    }
  }
}
```
```

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

### datasaver Environment Variable Configuration

Source: https://github.com/localrivet/datasaver/blob/main/README.md

Lists and explains environment variables used to configure datasaver. Covers database connection details for PostgreSQL and SQLite, backup scheduling, storage options (local and S3), retention policies, compression, and monitoring settings.

```bash
# Database Type (postgres or sqlite)
DATASAVER_DB_TYPE=postgres         # postgres (default) or sqlite

# PostgreSQL Connection
DATASAVER_DB_HOST=postgres
DATASAVER_DB_PORT=5432
DATASAVER_DB_NAME=mydb
DATASAVER_DB_USER=postgres
DATASAVER_DB_PASSWORD=secret

# Or use connection string (PostgreSQL)
DATASAVER_DATABASE_URL=postgres://user:pass@host:5432/dbname

# SQLite Configuration
DATASAVER_DB_TYPE=sqlite
DATASAVER_DB_PATH=/data/myapp.db   # Path to SQLite database file

# Backup Schedule (cron format)
DATASAVER_SCHEDULE="0 2 * * *"

# Storage
DATASAVER_STORAGE_BACKEND=local    # local or s3
DATASAVER_STORAGE_PATH=/backups    # for local storage

# S3 Storage
DATASAVER_S3_BUCKET=my-backups
DATASAVER_S3_ENDPOINT=s3.amazonaws.com
DATASAVER_S3_REGION=us-east-1
DATASAVER_S3_ACCESS_KEY=xxx
DATASAVER_S3_SECRET_KEY=xxx
DATASAVER_S3_USE_SSL=true

# Retention Policy
DATASAVER_KEEP_DAILY=7
DATASAVER_KEEP_WEEKLY=4
DATASAVER_KEEP_MONTHLY=6
DATASAVER_MAX_AGE_DAYS=90

# Compression
DATASAVER_COMPRESSION=gzip         # gzip or none

# Monitoring
DATASAVER_METRICS_PORT=9090
DATASAVER_HEALTH_PORT=8080
DATASAVER_WEBHOOK_URL=https://hooks.example.com/backup
DATASAVER_ALERT_AFTER_HOURS=26
```

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

### Datasaver Configuration - Environment Variables (Bash)

Source: https://github.com/localrivet/datasaver/blob/main/prd.md

Defines the environment variables used to configure the datasaver application. These variables cover database connection details, backup scheduling, storage backend settings (local or S3), retention policies, compression methods, and monitoring parameters. Some variables like DATASAVER_STORAGE_BACKEND and DATASAVER_S3_ENDPOINT use specific values to indicate the chosen configuration.

```bash
# Database Connection
DATASAVER_DB_HOST=postgres
DATASAVER_DB_PORT=5432
DATASAVER_DB_NAME=mydb
DATASAVER_DB_USER=postgres
DATASAVER_DB_PASSWORD=secret

# Or use connection string
DATASAVER_DATABASE_URL=postgres://user:pass@host:5432/dbname

# Backup Schedule
DATASAVER_SCHEDULE="0 2 * * *"  # 2 AM daily

# Storage
DATASAVER_STORAGE_BACKEND=s3    # local|s3
DATASAVER_STORAGE_PATH=/backups # for local
DATASAVER_S3_BUCKET=my-backups
DATASAVER_S3_ENDPOINT=s3.amazonaws.com
DATASAVER_S3_REGION=us-east-1
DATASAVER_S3_ACCESS_KEY=xxx
DATASAVER_S3_SECRET_KEY=xxx

# Retention
DATASAVER_KEEP_DAILY=7
DATASAVER_KEEP_WEEKLY=4
DATASAVER_KEEP_MONTHLY=6
DATASAVER_MAX_AGE_DAYS=90

# Compression
DATASAVER_COMPRESSION=gzip      # gzip|zstd|none

# Monitoring
DATASAVER_METRICS_PORT=9090
DATASAVER_WEBHOOK_URL=https://hooks.example.com/backup
DATASAVER_ALERT_AFTER_HOURS=26  # Alert if no backup in 26 hours
```

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

### Running DataSaver Daemon with YAML Configuration

Source: https://github.com/localrivet/datasaver/blob/main/docs/configuration.md

This command demonstrates how to run the DataSaver daemon while specifying a custom configuration file path. The `-c` flag is used to point to the location of the `config.yaml` file.

```bash
datasaver daemon -c /path/to/config.yaml

```

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

### datasaver CLI Commands for Backup Operations

Source: https://github.com/localrivet/datasaver/blob/main/README.md

Demonstrates common datasaver CLI commands for performing backups, listing existing backups, restoring from a specific backup, and checking the system's health. These commands can be executed directly from the terminal.

```bash
# Run as daemon
datasaver daemon

# Perform immediate backup
datasaver backup

# List backups
datasaver list

# Restore from backup
datasaver restore backup_20240111_0200

# Check health
datasaver health
```

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

### Comprehensive Datasaver YAML Configuration

Source: https://context7.com/localrivet/datasaver/llms.txt

Provides a detailed YAML configuration file for Datasaver, covering database connection (PostgreSQL and SQLite), storage options (local and S3), backup scheduling, compression, retention policies, backup verification, and monitoring settings. Environment variable expansion is supported for sensitive parameters like passwords.

```yaml
# config.yaml
database:
  type: postgres
  host: localhost
  port: 5432
  name: myapp
  user: backup_user
  password: ${DB_PASSWORD}  # Environment variable expansion supported

storage:
  backend: s3
  path: ./backups  # Used for local backend
  s3:
    bucket: my-backups
    endpoint: s3.amazonaws.com
    region: us-west-2
    access_key: ${AWS_ACCESS_KEY_ID}
    secret_key: ${AWS_SECRET_ACCESS_KEY}
    use_ssl: true

schedule: "0 */6 * * *"  # Every 6 hours
compression: gzip  # gzip, zstd, or none

retention:
  daily: 7
  weekly: 4
  monthly: 12
  max_age_days: 90

backup:
  verify_after_backup: true  # Verify backup integrity after creation
  verify_checksum: true      # Verify checksum on restore

monitoring:
  health_port: 8080
  metrics_port: 9090
  webhook_url: https://hooks.slack.com/services/xxx/yyy/zzz
  alert_after_hours: 26
```

```yaml
# config-sqlite.yaml
database:
  type: sqlite
  path: /data/myapp.db

storage:
  backend: local
  path: /backups

schedule: "0 3 * * *"
compression: gzip

retention:
  daily: 7
  weekly: 4
  monthly: 6
```

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

### Datasaver CLI - Health Check (Bash)

Source: https://github.com/localrivet/datasaver/blob/main/prd.md

Shows how to perform a health check on the datasaver backup system using the CLI. The `datasaver health` command provides a summary of the system's status, including the last and next backup times, total number of backups, and current storage utilization. A sample output is provided.

```bash
# Check backup system health
datasaver health
# Output:
# Status: healthy
# Last backup: 2024-01-11 02:00:15 (success)
# Next backup: 2024-01-12 02:00:00
# Total backups: 23
# Storage used: 2.8 GB
```

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

### Claude Desktop MCP Configuration (JSON)

Source: https://context7.com/localrivet/datasaver/llms.txt

JSON configuration file for Claude Desktop to connect to the datasaver MCP endpoint. It specifies the endpoint URL and includes the necessary Bearer token for authentication.

```json
// claude_desktop_config.json
{
  "mcpServers": {
    "datasaver": {
      "url": "http://localhost:8080/mcp",
      "headers": {
        "Authorization": "Bearer your-secure-api-key"
      }
    }
  }
}
```

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

### Datasaver Docker Compose Integration

Source: https://github.com/localrivet/datasaver/blob/main/prd.md

Provides a docker-compose.yml file for deploying the datasaver application alongside a PostgreSQL database. It defines two services: 'postgres' for the database and 'datasaver' for the application itself. The 'datasaver' service is configured with environment variables and volumes, and it depends on the 'postgres' service to ensure proper startup order.

```yaml
version: "3.8"

services:
  postgres:
    image: postgres:16-alpine
    environment:
      POSTGRES_DB: mydb
      POSTGRES_USER: postgres
      POSTGRES_PASSWORD: secret
    volumes:
      - postgres_data:/var/lib/postgresql/data
    networks:
      - app_network

  datasaver:
    image: ghcr.io/yourusername/datasaver:latest
    environment:
      DATASAVER_DB_HOST: postgres
      DATASAVER_DB_NAME: mydb
      DATASAVER_DB_USER: postgres
      DATASAVER_DB_PASSWORD: secret
      DATASAVER_SCHEDULE: "0 2 * * *"
      DATASAVER_STORAGE_BACKEND: local
      DATASAVER_STORAGE_PATH: /backups
      DATASAVER_KEEP_DAILY: 7
      DATASAVER_KEEP_WEEKLY: 4
      DATASAVER_KEEP_MONTHLY: 6
    volumes:
      - backup_data:/backups
    depends_on:
      - postgres
    networks:
      - app_network
    restart: unless-stopped

volumes:
  postgres_data:
  backup_data:

networks:
  app_network:
```

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

### backup_now Tool

Source: https://github.com/localrivet/datasaver/blob/main/docs/mcp-integration.md

Triggers an immediate backup of the data.

```APIDOC
## POST /mcp/tools/backup_now

### Description
Triggers an immediate backup.

### Method
POST

### Endpoint
`/mcp/tools/backup_now`

### Parameters
#### Query Parameters
- **limit** (integer) - Optional - The maximum number of backups to list.

#### Request Body
- **name** (string) - Required - The name of the tool, must be "backup_now".
- **arguments** (object) - Required - An empty object for this tool.

### Request Example
```json
{
  "name": "backup_now",
  "arguments": {}
}
```

### Response
#### Success Response (200)
- **result** (object) - The result of the backup operation.
- **error** (string) - An error message if the operation failed.

#### Response Example
```json
{
  "result": {
    "backup_id": "20240115-020000"
  },
  "error": null
}
```
```

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

### list_backups Tool

Source: https://github.com/localrivet/datasaver/blob/main/docs/mcp-integration.md

Lists all available backups, with an optional limit.

```APIDOC
## POST /mcp/tools/list_backups

### Description
Lists all available backups.

### Method
POST

### Endpoint
`/mcp/tools/list_backups`

### Parameters
#### Request Body
- **name** (string) - Required - The name of the tool, must be "list_backups".
- **arguments** (object) - Required - Contains arguments for the tool.
  - **limit** (integer) - Optional - The maximum number of backups to list.

### Request Example
```json
{
  "name": "list_backups",
  "arguments": {
    "limit": 10
  }
}
```

### Response
#### Success Response (200)
- **result** (object) - The result of the list operation.
  - **backups** (array) - A list of backup objects.
    - **backup_id** (string) - The ID of the backup.
    - **timestamp** (string) - The timestamp of the backup.
- **error** (string) - An error message if the operation failed.

#### Response Example
```json
{
  "result": {
    "backups": [
      {
        "backup_id": "20240115-020000",
        "timestamp": "2024-01-15T02:00:00Z"
      }
    ]
  },
  "error": null
}
```
```

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

### MCP Tool: backup_now (JSON)

Source: https://github.com/localrivet/datasaver/blob/main/docs/mcp-integration.md

Represents the JSON payload for the backup_now tool, which triggers an immediate backup. This tool takes no arguments.

```json
{
  "name": "backup_now",
  "arguments": {}
}
```

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

### Prometheus Metrics Endpoint

Source: https://context7.com/localrivet/datasaver/llms.txt

This endpoint exposes Prometheus-compatible metrics for monitoring and observability of the datasaver service, including backup duration, size, success rates, and storage utilization.

```APIDOC
## GET /metrics

### Description
Exposes Prometheus metrics for observability.

### Method
GET

### Endpoint
/metrics

### Parameters
None

### Request Example
```bash
curl http://localhost:9090/metrics
```

### Response
#### Success Response (200)
Returns a text-based response containing Prometheus metrics. Key metrics include:
- **datasaver_backup_duration_seconds**: Histogram of backup duration.
- **datasaver_backup_size_bytes**: Gauge of the last backup size.
- **datasaver_backups_total**: Counter of total backup attempts.
- **datasaver_backup_failures_total**: Counter of failed backups.
- **datasaver_last_backup_timestamp**: Gauge of the last backup time.
- **datasaver_last_backup_success**: Gauge (1 for success, 0 for failure).
- **datasaver_storage_used_bytes**: Gauge of total storage used.

#### Response Example
```text
# HELP datasaver_backup_duration_seconds Histogram of backup duration
# TYPE datasaver_backup_duration_seconds histogram
datasaver_backup_duration_seconds_bucket{le="0.1"} 0
datasaver_backup_duration_seconds_bucket{le="0.5"} 1
datasaver_backup_duration_seconds_bucket{le="1"} 2
datasaver_backup_duration_seconds_bucket{le="+Inf"} 2
datasaver_backup_duration_seconds_sum 1.2345
datasaver_backup_duration_seconds_count 2
# HELP datasaver_backup_size_bytes Gauge of last backup size
# TYPE datasaver_backup_size_bytes gauge
datasaver_backup_size_bytes 10485760
# HELP datasaver_backups_total Counter of total backup attempts
# TYPE datasaver_backups_total counter
datasaver_backups_total 5
# HELP datasaver_backup_failures_total Counter of failed backups
# TYPE datasaver_backup_failures_total counter
datasaver_backup_failures_total 1
# HELP datasaver_last_backup_timestamp Gauge of last backup time
# TYPE datasaver_last_backup_timestamp gauge
datasaver_last_backup_timestamp 1673779215
# HELP datasaver_last_backup_success Gauge (1=success, 0=failure)
# TYPE datasaver_last_backup_success gauge
datasaver_last_backup_success 1
# HELP datasaver_storage_used_bytes Gauge of total storage used
# TYPE datasaver_storage_used_bytes gauge
datasaver_storage_used_bytes 524288000
```
```

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

### Querying Datasaver Health Endpoint

Source: https://context7.com/localrivet/datasaver/llms.txt

This command queries the Datasaver health endpoint to retrieve the current status of the backup system. The response includes the system's status ('healthy' or 'unhealthy'), the timestamp of the last backup, and the scheduled time for the next backup. An unhealthy response may also include the last error encountered.

```bash
# Query health endpoint
curl http://localhost:8080/health

# Expected response:
# status: healthy
# last_backup: 2024-01-15T02:00:15Z
# next_backup: 2024-01-16T02:00:00Z

# Unhealthy response (503 status):
# status: unhealthy
# last_backup: 2024-01-15T02:00:15Z
# last_error: database dump failed: connection refused
# next_backup: 2024-01-16T02:00:00Z
```

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

### verify_backup Tool

Source: https://github.com/localrivet/datasaver/blob/main/docs/mcp-integration.md

Verifies the integrity of a specified backup.

```APIDOC
## POST /mcp/tools/verify_backup

### Description
Verifies backup integrity.

### Method
POST

### Endpoint
`/mcp/tools/verify_backup`

### Parameters
#### Request Body
- **name** (string) - Required - The name of the tool, must be "verify_backup".
- **arguments** (object) - Required - Contains arguments for the tool.
  - **backup_id** (string) - Required - The ID of the backup to verify.

### Request Example
```json
{
  "name": "verify_backup",
  "arguments": {
    "backup_id": "20240115-020000"
  }
}
```

### Response
#### Success Response (200)
- **result** (object) - The result of the verification operation.
  - **is_valid** (boolean) - True if the backup is valid, false otherwise.
- **error** (string) - An error message if the operation failed.

#### Response Example
```json
{
  "result": {
    "is_valid": true
  },
  "error": null
}
```
```

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

### get_status Tool

Source: https://github.com/localrivet/datasaver/blob/main/docs/mcp-integration.md

Retrieves the current status of the backup system.

```APIDOC
## POST /mcp/tools/get_status

### Description
Gets the current backup system status.

### Method
POST

### Endpoint
`/mcp/tools/get_status`

### Parameters
#### Request Body
- **name** (string) - Required - The name of the tool, must be "get_status".
- **arguments** (object) - Required - An empty object for this tool.

### Request Example
```json
{
  "name": "get_status",
  "arguments": {}
}
```

### Response
#### Success Response (200)
- **result** (object) - The status of the backup system.
  - **status** (string) - The current status (e.g., "idle", "running", "error").
  - **message** (string) - A detailed message about the status.
- **error** (string) - An error message if the operation failed.

#### Response Example
```json
{
  "result": {
    "status": "idle",
    "message": "Backup system is ready."
  },
  "error": null
}
```
```

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

### Backup Metadata JSON Format

Source: https://github.com/localrivet/datasaver/blob/main/prd.md

Defines the structure of the metadata JSON file included in each backup. It contains information about the backup's ID, timestamp, database details, backup method, compression, size, duration, checksum, file list, and retention policy.

```json
{
  "id": "backup_20240111_020015",
  "timestamp": "2024-01-11T02:00:15Z",
  "type": "daily",
  "database": {
    "name": "mydb",
    "host": "postgres",
    "version": "16.1"
  },
  "backup": {
    "method": "pg_dump",
    "format": "custom",
    "compression": "gzip",
    "size_bytes": 131621888,
    "compressed_size_bytes": 45678912,
    "duration_seconds": 12.5,
    "checksum": "sha256:abcd1234..."
  },
  "files": [
    "backup_20240111_020015.dump.gz",
    "backup_20240111_020015.meta.json"
  ],
  "retention": {
    "keep_until": "2024-02-11T02:00:15Z",
    "policy": "daily"
  }
}
```