GK Healter

🇹🇷 Türkçe (Turkish)  |  🇬🇧 English

GK Healter is a professional system maintenance and health-monitoring utility designed primarily for Pardus and Debian-based Linux distributions. It emphasizes safety and efficiency, providing users with reliable disk space recovery, error detection, and proactive maintenance capabilities without compromising system stability.

🏆 Developed for TEKNOFEST 2026 — Pardus Bug Catching and Suggestion Competition (Development Category).

Developed by Egehan KAHRAMAN and Mustafa GÖKPINAR — GK Developers.

Key Features

Pardus-Specific Diagnostics

• Pardus Repository Health Check: Validates APT sources, detects broken/held packages, checks Pardus-specific services (pardus-*, eta-*).
• Pardus Version Detection: Identifies Pardus release info and provides distribution-aware recommendations.
• Broken Package Detection: Uses dpkg --audit and apt-get check for Debian/Pardus-native package integrity.

System Maintenance

• Package Management: Clean APT cache, autoremove orphan packages, fix broken dependencies via polkit-authenticated actions.
• System Log Cleanup: Remove redundant logs, vacuum systemd journal, clean old coredumps.
• Application Hygiene: Clear browser caches, thumbnail galleries, and user-specific temporary files.
• Safety-First Approach: Whitelist-based deletion prevents accidental removal of critical system files.

Monitoring & Intelligence

• Real-Time Health Score: CPU, RAM, and disk usage monitoring with a composite health score (0–100).
• Hybrid AI Analysis: Offline rule-based LocalAnalysisEngine always available; optional cloud enrichment via Gemini/OpenAI APIs.
• Smart Recommendations: Rule-based engine generates actionable suggestions based on system metrics.
• Service Analyzer: Detects failed systemd services and slow-starting units.
• Log Analyzer: Identifies critical/error-level journal entries with severity classification.

Security Audit

• World-Writable File Detection: Scans system directories for insecure permissions.
• SUID/SGID Binary Audit: Identifies unexpected set-uid binaries against a known whitelist.
• Sudoers Risk Analysis: Flags dangerous NOPASSWD: ALL entries.
• SSH Hardening Check: Validates sshd_config against security best practices.
• Unattended-Upgrades Monitoring: Verifies automatic security updates are enabled.
• Failed Login Tracking: Summarizes authentication failures from journal.

Automation

• Intelligent Auto-Maintenance: Scheduled cleaning based on idle time, disk thresholds, and power status.
• Cleaning History: Comprehensive tracking of all operations with timestamps and space recovered.

User Experience

• Native GTK 3 Interface: Modern, responsive design that respects system themes and dark mode.
• Multi-Language Support: Turkish and English with extensible JSON-based i18n system.
• Report Export: Generate TXT, HTML, and JSON system analysis reports for documentation and demo purposes.
• Dedicated Security Page: Separate tab with Pardus verification data, colour-coded findings, and one-click export.

Screenshot

Technology Stack

• Language: Python 3
• GUI Toolkit: GTK 3 (PyGObject)
• Build System: Meson / Make
• Testing: pytest (298 tests, %75+ coverage) with CI via GitHub Actions
• Packaging: Flatpak, Debian (.deb), Arch (PKGBUILD), RPM (.spec)
• Privilege Escalation: Polkit (pkexec) with custom policy file

Architecture

src/
├── main.py                  # Application entry point
├── ui.py                    # GTK UI controller (Builder pattern)
├── cleaner.py               # Safety-first cleaning engine
├── health_engine.py         # Real-time system health monitoring
├── pardus_analyzer.py       # Pardus/Debian-specific diagnostics
├── security_scanner.py      # System security audit engine
├── pardus_verifier.py       # Pardus identity verification & evidence
├── report_exporter.py       # TXT / HTML / JSON report generator
├── distro_manager.py        # Multi-distro package manager abstraction
├── disk_analyzer.py         # Large file discovery
├── log_analyzer.py          # Journal error analysis
├── service_analyzer.py      # Systemd service health
├── recommendation_engine.py # Rule-based system recommendations
├── ai_engine.py             # Hybrid AI: local analysis + cloud enrichment
├── auto_maintenance_manager.py # Scheduled maintenance logic
├── settings_manager.py      # Persistent configuration
├── history_manager.py       # Cleaning history tracking
├── i18n_manager.py          # Internationalization (JSON-based)
├── logger.py                # Centralized logging (rotating files)
└── utils.py                 # Shared utility functions

Installation

Pardus / Debian / Ubuntu (.deb) — Recommended

cd gk-healter
make deb
sudo dpkg -i gk-healter_0.1.6_all.deb
sudo apt-get install -f  # Fix any missing dependencies

Flatpak (any distro)

flatpak install flathub io.github.gkdevelopers.GKHealter
flatpak run io.github.gkdevelopers.GKHealter

Arch Linux (AUR / PKGBUILD)

cd packaging/arch
makepkg -si

Fedora / openSUSE (RPM)

rpmbuild -ba packaging/rpm/gk-healter.spec

Generic install (any distro)

cd gk-healter
sudo make install
# To remove:
sudo make uninstall

Build from Source (Meson)

git clone https://github.com/GK-Developers/GK-Healter.git
cd GK-Healter/gk-healter

meson setup _build
meson compile -C _build
sudo meson install -C _build

Build Dependencies

Dependency	Pardus/Debian	Arch	Fedora
Python 3	python3	python	python3
PyGObject	python3-gi	python-gobject	python3-gobject
GTK 3	gir1.2-gtk-3.0	gtk3	gtk3
psutil	python3-psutil	python-psutil	python3-psutil
Polkit	policykit-1	polkit	polkit
Meson	meson	meson	meson

Running Tests

pip install pytest pytest-cov
pytest -v --cov=src --cov-report=term-missing

298 tests covering 17 modules across 18 test files.

Docker-based Pardus 25 test environment

For competition demos and offline verification on vulnerable / corrupted Pardus systems, there is a dedicated Docker-based test harness:

• Location: tests/docker/
• Base image: pardus/yirmibes:latest (or pardus/etap)
• Artifacts: TXT/HTML/JSON reports + *.manifest.json stored under artifacts/

Typical workflow:

# build once on a connected machine
docker build -t gk-healter-test:pardus25 -f tests/docker/Dockerfile .
docker save gk-healter-test:pardus25 -o gk-healter-test_pardus25.tar

# transfer tar, then on the (offline) Pardus box
docker load -i gk-healter-test_pardus25.tar
docker run --rm -it --privileged -v /dev:/dev -v "$(pwd)":/workspace gk-healter-test:pardus25

# inside the container
bash tests/docker/run_all_scenarios.sh
ls -la /workspace/artifacts

Each scenario (baseline, bloat, corruption, SUID backdoor, repo breakage, pseudo-malware, etc.) produces its own *-<tag>.txt/html/json/manifest.json bundle. A consolidated human-readable summary of one full run lives at GKHealter_DockerSecurityEvaluation_2026-03-11.md.

Packaging

Format	File	Location
Flatpak	flathub_submission.yml	gk-healter/
Arch Linux	PKGBUILD	gk-healter/packaging/arch/
RPM	gk-healter.spec	gk-healter/packaging/rpm/
Debian/Pardus	debian/control	gk-healter/debian/

Contributing

See CONTRIBUTING.md for development setup and contribution guidelines.

License

This project is licensed under the MIT License. See the LICENSE file for details.

Project Links

• Homepage: https://github.com/GK-Developers/GK-Healter
• Bug Tracker: https://github.com/GK-Developers/GK-Healter/issues
• Source Code: https://github.com/GK-Developers/GK-Healter