Local-First Storage Intelligence and Developer Storage Analyzer for macOS, Windows, and Linux.
Overview | Key Features | Comparison | Architecture | Downloads | Building from Source | Supported Ecosystems | Security & Privacy | Contributing | License
Modern software development consumes significant storage. A single active software engineer easily accumulates 50 to 150 GB of redundant build artifacts, nested node_modules, dormant Python virtual environments, compiler object files, Docker build caches, and stale IDE indexing databases.
Existing storage utilities present substantial limitations:
- Legacy utilities (such as WinDirStat) rely on single-threaded architectures designed decades ago, leading to long scan times, high memory overhead, and unresponsiveness on multi-terabyte SSDs.
- Commercial cleanup applications frequently require recurring subscriptions, install persistent background helper daemons, collect remote telemetry, and lack awareness of software development tooling.
- Standard operating system analyzers display generic categories such as "System Data" or "Documents" without granular breakdown of user dotfiles or project directories.
DiskAtlas is an open-source, local-first storage intelligence application designed specifically for developers, systems engineers, and power users. Built on a multi-threaded Rust backend and Tauri 2 shell with React 19 and D3.js, DiskAtlas indexes drives at hardware speeds, accurately classifies project dependencies, visualizes storage distribution, and recommends reclaimable space through transparent, user-controlled workflows.
- Multi-threaded traversal engine written in Rust, delivering throughput exceeding 50,000 files per second on modern NVMe and APFS drives.
- Native filesystem boundary awareness: on macOS APFS, automatically unifies container partitions to prevent duplicate counting between the sealed system volume (
/) and the data volume (/System/Volumes/Data). - Cycle-proof traversal: detects and bypasses directory loops, Windows NTFS junctions (
IO_REPARSE_TAG_MOUNT_POINT), and special operating system mount points (/dev,/proc,/sys). - Dynamic bottom-up directory aggregation: recursively accumulates child directory sizes so every folder reflects its accurate total storage weight.
- Non-blocking controls: support for real-time pausing, resuming, and safe cancellation without database corruption.
Identifies software development artifacts across major toolchains and distinguishes user source code from regenerable build outputs:
- Node.js & JavaScript: Locates
node_modules,.next/cache,.nuxt,.output,.vite,.turbo,.parcel-cache, npm global caches, and pnpm stores. Extracts package identities directly frompackage.json. - Python: Detects
.venv,venv,__pycache__,.pytest_cache,.mypy_cache,.ruff_cache, pip wheel caches, and Conda/Poetry environments. - Rust: Discovers
target/directories,.cargo/registry, and.cargo/gitcache stores. - Docker: Inspects images, containers, volumes, and build cache via local Docker engine APIs without directly manipulating internal storage files.
- Java & Android: Indexes
.m2/repository,.gradle/caches, Android SDK platforms, emulator system images, and JDK installations. - C & C++: Identifies
cmake-build-debug,cmake-build-release,.vs/, and compiler object files. - IDEs: Identifies cache and indexing footprints from VS Code, Cursor, JetBrains, Xcode DerivedData, and Antigravity.
- D3 Squarified Treemap: Hierarchical, area-proportional rectangle layout with category-based styling, breadcrumb navigation, zoom, and live drilldown.
- D3 Radial Sunburst: Multi-level concentric partition chart illustrating directory depth and structural distribution.
- Storage Analytics: Aggregated views for category distribution, file extension breakdowns, largest individual consumers, and historical volume trends.
- Staged Duplicate Finder: 4-stage pipeline (Size match -> Metadata comparison -> Block header check -> Full SHA-256 cryptographic hash match). Distinguishes identical files from visually similar media without relying on identical file names.
- Large Files Engine: Configurable size thresholds, extension filtering, path whitelists, and fast column sorting.
- Hidden & System Storage Breakdown: Distinguishes user hidden items (dotfiles and user-hidden attributes) from critical system-protected operating system components.
- Forgotten Files: Identifies stale installers (
.dmg,.pkg,.exe,.msi), old archives (.zip,.tar.gz), and inactive project artifacts based on access and modification metrics. - Unknown Storage Discovery: Categorizes unclassified binary blobs and isolated subtrees to ensure zero unmapped drive space.
- Zero Automatic Deletions: Every deletion requires explicit selection, review, and confirmation.
- System Protection Rules: Automatically blocks deletion requests targeting critical boot components, drivers, kernel paging files, and sealed operating system snapshots.
- Cloud-Sync Safeguards: Identifies directories synchronized with OneDrive, Google Drive, Dropbox, or iCloud to warn users regarding remote deletion propagation.
- Reversible Workflow: Directs deletions to the native operating system Trash or Windows Recycle Bin whenever supported.
- Immutable Audit Trail: Logs every cleanup operation into an embedded SQLite audit table, recording target paths, timestamps, execution methods, and status outcomes.
- Zero Remote Telemetry: Contains no third-party tracking libraries, analytics SDKs, or background telemetry.
- 100% Offline Operation: Functions fully without internet access.
- Embedded SQLite Engine: Stores all indexed metadata locally using SQLite with Write-Ahead Logging (WAL) and memory-mapped I/O, maintaining an operational memory footprint under 50 MB.
| Feature | DiskAtlas | WinDirStat | DaisyDisk | CleanMyMac | Baobab |
|---|---|---|---|---|---|
| License | MIT (Open Source) | GPL (Open Source) | Proprietary | Proprietary | GPL (Open Source) |
| Platform Support | macOS, Windows, Linux | Windows only | macOS only | macOS only | Linux only |
| Pricing | Free | Free | $9.99 | $39.95 / year | Free |
| Developer Intelligence | Yes | No | No | Partial | No |
| Engine Architecture | Multi-threaded Rust | Single-threaded C++ | Native Swift | Proprietary | C / GTK |
| File Traversal Speed | >50,000 files/sec | ~5,000 files/sec | High | Moderate | Moderate |
| APFS Container Unification | Yes | N/A | Yes | Partial | N/A |
| Interactive Visualizations | Treemap & Sunburst | Treemap only | Sunburst only | Custom gauges | Treemap & Rings |
| Reversible Cleanup | Recycle Bin / Trash | Permanent deletion | Permanent deletion | Custom script | Trash |
| Telemetry | Zero (Strictly Local) | Zero | Unknown | Included | Zero |
+-----------------------------------------------------------------------+
| DiskAtlas Shell |
| React 19 * TypeScript * Tailwind CSS * D3.js Charts |
+-----------------------------------▲-----------------------------------+
| Tauri 2 IPC (Type-Safe RPC)
+-----------------------------------▼-----------------------------------+
| Rust Core Engine |
| * Universal Multi-Threaded Scanner (Bounded Stack, Cycle-Proof) |
| * Filesystem Adapters (APFS Single Container, Windows NTFS Win32) |
| * 16-Category Deterministic Classification Engine |
| * Developer Toolchain Analyzers (Cargo, npm, pip, Docker, Xcode) |
| * Reversible Trash & Safety Barrier Service |
+-----------------------------------▲-----------------------------------+
| rusqlite (Bundled SQLite 3)
+-----------------------------------▼-----------------------------------+
| Embedded Database (WAL) |
| * Indexed Files & Recursive Directory Trees |
| * Point-in-Time Storage Snapshots & Differential Diffs |
| * Cryptographic Duplicate Hashes & Audit Events |
+-----------------------------------------------------------------------+
Official release packages are published on the GitHub Releases page.
- macOS: 11.0 (Big Sur) and newer. Universal binary supports Apple Silicon (M1/M2/M3/M4) and Intel (x86_64). Available as
.dmg. - Windows: Windows 10 and Windows 11 (64-bit and ARM64). Available as an NSIS standalone installer (
.exe) and enterprise.msipackage. - Linux: Major distributions including Ubuntu, Debian, Fedora, and Arch. Available as
.AppImageand.deb.
# macOS (Homebrew Cask)
brew install --cask diskatlas
# Windows (Winget)
winget install Harshit17x.DiskAtlas
# Windows (Scoop)
scoop bucket add extras
scoop install diskatlas
# Cargo (crates.io source build)
cargo install diskatlas| Dependency | Minimum Version | Notes |
|---|---|---|
| Node.js | v20 LTS | Required for frontend tooling |
| Rust & Cargo | v1.80+ | Install via rustup.rs |
| Platform Tools | Platform dependent | Windows: Visual Studio C++ Build Tools macOS: Xcode Command Line Tools ( xcode-select --install)Linux: libwebkit2gtk-4.1-dev, build-essential, libssl-dev |
-
Clone the repository:
git clone https://github.com/Harshit17x/Diskatlas.git cd Diskatlas -
Install frontend dependencies:
npm install
-
Run the automated test suite:
cargo test --manifest-path src-tauri/Cargo.toml -
Launch the application in development mode with hot reloading:
npm run tauri dev
-
Compile a production release bundle for your platform:
npm run build npm run tauri build
Production artifacts will be placed in
src-tauri/target/release/bundle/.
DiskAtlas identifies project roots using marker files and classifies internal dependencies:
| Toolchain / Ecosystem | Marker File | Detected Artifacts & Reclaimable Locations |
|---|---|---|
| Node.js / JS / TS | package.json, package-lock.json, pnpm-lock.yaml, yarn.lock, bun.lockb |
node_modules, .next/cache, .nuxt, .output, .vite, .turbo, .parcel-cache, coverage, npm/pnpm caches |
| Python | pyproject.toml, setup.py, Pipfile, requirements.txt |
.venv, venv, __pycache__, .pytest_cache, .mypy_cache, .ruff_cache, pip wheels cache, Conda/Poetry envs |
| Rust | Cargo.toml, Cargo.lock |
target/, .cargo/registry, .cargo/git |
| Docker | Local Docker Engine socket / daemon | Unused container images, stopped containers, dangling volumes, and build cache via official APIs |
| Java & Kotlin | pom.xml, build.gradle, build.gradle.kts |
target/, build/, .m2/repository, .gradle/caches |
| C & C++ | CMakeLists.txt, Makefile, *.vcxproj |
cmake-build-debug, cmake-build-release, .vs/, intermediate object files (*.o, *.obj) |
| Mobile & Apple | *.xcodeproj, *.xcworkspace, Podfile |
Xcode DerivedData, CoreSimulator runtimes, Android SDK platforms, emulator system images |
| Version Control & IDEs | .git, .vscode, .idea |
.git/objects packfiles, Git LFS store, language server index caches, telemetry buffers |
DiskAtlas is engineered with a strict security boundary:
- Read-Only Scanner Handles: All discovery routines open filesystem nodes with read-only metadata permissions. The scanner contains zero destructive primitives.
- Explicit Two-Step Deletion: Destructive cleanup actions require selecting items individually, reviewing total size and path previews, and confirming intent.
- System Directory Guard: Paths belonging to operating system kernels, bootloaders, drivers, or system frameworks are blocked from cleanup selection.
- Isolated Network Boundary: The binary initiates zero outbound network requests. All analytics and storage diagnostics are processed and stored locally.
Refer to docs/SECURITY_AUDIT.md for the complete security review and threat model assessment.
Contributions are welcome. Please read CONTRIBUTING.md for environment setup and pull request procedures.
- Ecosystem Parsers: Adding detection rules for additional frameworks (e.g., Elixir
_build, Go module cache, Dart/Flutter pub cache). - Filesystem Optimizations: Profiling scanner performance across unique storage topologies.
- Visualizations: Enhancing accessibility, color contrast modes, and chart interactions.
- Localization: Translating interface strings into additional languages.
Please ensure all automated tests pass before submitting pull requests:
cargo test --manifest-path src-tauri/Cargo.toml
npm run buildDiskAtlas is open-source software licensed under the MIT License.