toolkitbox/all repository overview

⁠Toolkitbox

A modular Docker image build system for DevOps/SysAdmin tools and clients. You can use a single image with all tools included, individual images per tool, or bundles grouped by kind of tool.

⁠Available Images

⁠Bundle Images

Image	Description	Components	Size	Docker Pull	Badges
all	All tools in one image	all components	340.9 MB	docker pull toolkitbox/all
cloud	Cloud Bundle	aws-cli, azure-cli, gcloud	281.4 MB	docker pull toolkitbox/cloud
databases	Databases Bundle	postgres, mysql, mariadb, mongo, redis, sqlite	67.7 MB	docker pull toolkitbox/databases

⁠Individual Tools

Tool	Versions	Size	Docker Pull	Badges
aws-cli	latest,v1	45.3 MB	docker pull toolkitbox/aws-cli
azure-cli	latest	132.5 MB	docker pull toolkitbox/azure-cli
gcloud	latest	152.1 MB	docker pull toolkitbox/gcloud
mariadb	latest,v10.3,v10.4	17.2 MB	docker pull toolkitbox/mariadb
mongo	latest,v3.6,v4.4	53.2 MB	docker pull toolkitbox/mongo
mysql	latest,v5.6,v5.7	17.2 MB	docker pull toolkitbox/mysql
postgres	latest,v10,v11,v9.6	8.7 MB	docker pull toolkitbox/postgres
redis	latest,v4,v5	9.9 MB	docker pull toolkitbox/redis
sqlite	latest,v3	8.2 MB	docker pull toolkitbox/sqlite
duckdb	v1.4.4	-	docker pull toolkitbox/duckdb

⁠Standalone Only (Not in Bundles)

Some tools require glibc and cannot be included in Alpine-based bundles:

⁠Overview

This system allows you to dynamically build a "mega image" containing multiple tools without manually editing Dockerfiles. Simply add a new component directory with an install.sh script, and the orchestrator automatically includes it in the build.

This repository builds an individual image for each component as well, allowing for flexible usage and an all-in-one image.

⁠Architecture

The Dynamic Injection pattern works as follows:

1. Component Discovery: The orchestrator scans the components/ directory for tools
2. Script Collection: Each component's install.sh is copied to a temporary build context
3. Dynamic Installation: The Dockerfile iterates through all collected scripts and executes them
4. Clean Build: Temporary files are removed to keep the final image lean

⁠Project Structure

.gitignore                   # Ignores .build-context directory
build-mega-image.sh          # Orchestrator script (main entry point)
build-single.sh              # Script to build individual component images
Dockerfile.bundle            # Generic Dockerfile for mega image
components/                  # Component library
├── postgres/
│   ├── install.sh           # Latest PostgreSQL client installer
│   └── v11-legacy/          # Legacy version (not included in bundle)
│       └── Dockerfile       # Standalone Dockerfile for PG 11
└── aws-cli/
└── install.sh           # AWS CLI installer

⁠Quick Start

⁠Building the all-in-one Image

./build-all-in-one.sh

The script will:

• Discover all components with install.sh files
• Create a temporary .build-context directory
• Copy and rename component installers (e.g., postgres.sh, aws-cli.sh)
• Build the Docker image tagged as ops-toolchain:latest

⁠Adding New Components

To add a new tool to the mega image:

1. Create a new directory under components/:

   mkdir components/my-tool
   

   Copy

2. Create an install.sh script:

   cat > components/my-tool/install.sh <<'EOF'
   #!/bin/bash
   set -e
   
   echo "Installing my-tool..."
   apk add --no-cache my-tool
   
   # Verify installation
   if command -v my-tool &> /dev/null; then
       echo "my-tool installed successfully"
   else
       echo "ERROR: my-tool installation failed"
       exit 1
   fi
   EOF
   

   Copy

3. Rebuild the all-in-one image:

   ./build-allinone.sh
   

   Copy

That's it! No Dockerfile editing required.

⁠Automated Docker Hub Description Sync

This repository includes an automated workflow that keeps Docker Hub image descriptions synchronized with component READMEs.

⁠Workflow: sync-dockerhub-descriptions.yml

Triggers:

• Automatically when components/**/README.md files are updated in the main branch
• Manually via GitHub Actions UI (workflow_dispatch)

Functionality:

1. Authenticates with Docker Hub API
2. Iterates through all components in components/ directory
3. Updates the Docker Hub description for each image using its README.md content

Requirements:

• DOCKER_TOKEN secret must be configured in repository settings
• Docker Hub repositories must exist before running the workflow

⁠Updating Component Descriptions

To update a Docker Hub image description:

1. Edit the component's README: components/{component}/README.md
2. Commit and push to the main branch
3. The workflow will automatically sync the description to Docker Hub

The workflow uses the entire README content as the "full description" on Docker Hub, preserving all markdown formatting.

⁠Contributing

When adding new components:

1. Follow the component guidelines above
2. Test the installation script in isolation first
3. Verify the individual and all-in-one image builds successfully
4. Update documentation if adding complex components
5. Create a comprehensive README.md in the component directory (it will be synced to Docker Hub)

⁠Author

Created and maintained by:

• Fermín Jiménez (@Rimander⁠)
• Pablo Pérez-Aradros (@pabpereza⁠)