Skip to content
VancoolaPublic

About

Adaptive traffic routing and network diagnostics for Windows.

Topics

Resources

Code of conduct

Contributing

Security policy

Stars

1 star

Watchers

0 watching

Forks

CatNet

CatNet is a Windows desktop utility for adaptive traffic routing, packet-processing profiles, and network diagnostics. It provides a focused interface for selecting, testing, and activating a network profile without requiring users to manage low-level command-line tools.

Important

CatNet is an early release. Review the current limitations before relying on it in a production environment.

Screenshots

Main window Automatic profile evaluation
CatNet main window CatNet evaluating an available traffic profile

Features

  • One-click activation and deactivation of the active traffic profile
  • Automatic evaluation of 21 bundled packet-processing strategies
  • Manual profile selection for advanced users
  • Connection diagnostics and service reachability checks
  • Windows startup and automatic activation options
  • System tray integration
  • Local preference storage and Windows-protected secret storage
  • English and Russian user interfaces

Current limitations

  • Windows is the only supported operating system.
  • Smart mode is the only complete routing mode in version 0.1.1.
  • Administrative permission is required when the native packet-processing engine is started.
  • Automatic rule updates and the rule editor are not implemented yet.

Architecture

CatNet uses Tauri 2 for the native application shell and Leptos for the WebAssembly user interface. The repository is split into two main packages:

  • src/ contains the Leptos frontend, application state, pages, and UI components.
  • src-tauri/ contains the native backend, settings, secure storage, diagnostics, tray integration, process management, and packet-processing engine.

The packaged application includes third-party networking components. Their licenses do not change the license of CatNet's original source code; see Third-party notices for details.

Building from source

Requirements

  • Windows 10 or later
  • The stable Rust toolchain
  • The wasm32-unknown-unknown Rust target
  • Trunk
  • Tauri CLI 2
  • The Windows development prerequisites required by Tauri, including Microsoft C++ Build Tools and WebView2

Install the Rust tools:

rustup target add wasm32-unknown-unknown
cargo install trunk --locked
cargo install tauri-cli --version "^2.0" --locked

Start the development application from an elevated terminal:

cargo tauri dev

Run the project checks:

cargo fmt --all -- --check
cargo check --workspace
cargo test --workspace
cargo clippy --workspace --all-targets -- -D warnings

Build release installers:

cargo tauri build

The repository also provides equivalent make verify and make release targets for environments with GNU Make installed.

Privacy and network access

CatNet stores non-sensitive preferences locally. Configuration links are stored through the Windows credential manager rather than in the regular settings file.

The diagnostics feature makes direct requests to its configured reachability targets to evaluate the current network profile. CatNet does not include telemetry or analytics in version 0.1.1.

Contributing

Bug reports and focused pull requests are welcome. Read CONTRIBUTING.md before submitting a change. Please report security-sensitive issues using the process in SECURITY.md.

Attribution and license

CatNet was originally created and developed by Vancoola.

CatNet's original source code is licensed under the Apache License 2.0. Redistributions must preserve the license and the attribution notices in NOTICE. The CatNet name and branding are not licensed for endorsement or for representing modified builds as official releases.

Bundled third-party components remain subject to their own licenses. See THIRD_PARTY_NOTICES.md.

About

Adaptive traffic routing and network diagnostics for Windows.

Topics

Resources

Code of conduct

Contributing

Security policy

Stars

1 star

Watchers

0 watching

Forks

Releases

Contributors

Languages