- Category
- Developer Tools
- Rank
- No. 1696Tools index
Previous survey · No. 1701 ·
- Pricing
- Open Source
- Type
- TOOL
- Builder
- yamadashy
- GitHub
- 3 stars
- Latest release
- v0.1.8
- Date
About
Disk usage visualizer for both AI and humans — categorizes directories (cache, build, logs) and renders an interactive treemap in your browser.
What it does
Duvis scans a chosen filesystem tree, measures each entry, and assigns practical categories such as build output, caches, archives, backups, and model stores. It serves interactive browser views for exploration, while terminal users and agents can consume trees, summaries, flat largest-entry lists, JSON, NDJSON, or compact TOON output.
Why it's ranked here
This is a credible diagnostic tool because it treats human inspection and machine consumption as equal use cases. Parallel scanning, explicit hardlink policy, composable filters, truncation metadata, and versioned structured output make the results practical beyond a pretty chart. Its refusal to delete anything also keeps investigation separate from cleanup.
What's good
The structured formats preserve scan context, skipped-item counts, hardlink policy, relative paths, and full subtree counts. Truncated views report both hidden entry count and hidden size. On Unix, allocated disk blocks and default hardlink deduplication align totals with common disk-usage expectations. Category explanations expose which naming rule matched.
Tradeoffs
Classification depends on names, and a classified directory passes its category to everything beneath it. That is useful grouping, but it cannot inspect content or prove that an item is disposable. Windows reports apparent size and does not deduplicate hardlinks. Display filters leave parent totals unchanged, which is honest but requires careful interpretation. The tool never performs cleanup.
How to use it well
Use it when developers or agents need to locate storage concentration, compare categories, or feed scan results into shell and data pipelines. Start with a summary or largest-entry list, then narrow by size, name, type, category, or modification age. Open the browser for spatial exploration. Keep a separate, deliberate process for deletion, retention policy, and automated cleanup.
Technical notes+
Cargo.toml defines a Rust 2021 crate whose default ui feature adds axum, tokio, and browser opening support; disabling default features removes that CLI surface. src/lib.rs exposes scanning, classification, entry, sorting, and hardlink types as a library API, while src/cli.rs owns binary dispatch. src/scan.rs delegates parallel walking and Unix allocation accounting to private submodules. src/wire.rs isolates serialization DTOs from domain types, and src/render.rs fixes the structured wire version at 2. src/filter.rs applies predicates after scanning. src/ui.rs describes an embedded HTTP server with /data.json, /rescan, and /reveal endpoints.
Observed
- License
- MIT
- Primary language
- Rust, edition 2021
- Packaging
- Published as a Cargo crate and installable with cargo install duvis
- Interfaces
- CLI, Rust library API, structured stdout formats, and local browser UI
- Structured formats
- JSON, NDJSON, and TOON, plus human-readable tree, summary, and largest-entry views
- Platform support
- macOS, Linux, and Windows
- UI stack
- React and d3 frontend served by an optional embedded axum and tokio server
Read from README.md, Cargo.toml, src/cli.rs, src/lib.rs, src/main.rs, src/ui.rs, src/scan.rs, src/wire.rs, src/entry.rs, src/filter.rs, src/render.rs, src/classify.rs, src/cli/app.rs, src/wire/ui.rs, src/cli/args.rs.
What it can do
Visualize disk usage
File system directories → Interactive treemap visualization
Categorize directories by type
Directory paths and contents → Categorized directory groups (cache, build, logs, etc.)
Render interactive treemap in browser
Disk usage data → Web-based interactive treemap interface
Analyze directory sizes
File system structure → Size measurements for directories and files
Identify storage-heavy directories
Disk usage analysis results → List of largest directories and their sizes
Tags
Tech Stack
Comments (0)
No comments yet
Editorially curated, with community endorsements as a secondary signal. Corrections welcome.