BLACKBOX

Local repository forensics for structure, hotspots, risks, dependencies, and suspicious code patterns.

Download: Windows · macOS · Linux · Run from source · Report an issue

What it does

• Repository structure and language mix
• Large files and code hotspots
• TODO/FIXME-style findings and suspicious-pattern signals
• Dependency and cycle signals
• Git repository context
• Exportable JSON reports

Download

Platform	File
Windows 10/11 (x64)	BLACKBOX-Windows-x64.exe — portable, no installer
macOS (Apple Silicon)	BLACKBOX-macOS-arm64.zip — unzip and move to Applications
Linux (x86_64)	BLACKBOX-Linux-x86_64.tar.gz — extract and run ./BLACKBOX

Each release is built from the tagged source by GitHub Actions and carries a SHA256SUMS.txt. The builds are not yet code-signed, so on first launch Windows SmartScreen may ask you to confirm ("More info" → "Run anyway"), and macOS may need you to Control-click the app and choose Open.

Status: beta. BLACKBOX does what this README describes and is covered by CI on Windows, macOS and Linux, but it is young: expect rough edges, and please report them.

Run from source

Requirements: Python 3.10+ with Tk support.

pyw blackbox_desktop.pyw

The application uses Python's standard library at runtime.

Build a standalone Windows executable

powershell -ExecutionPolicy Bypass -File .\build-windows.ps1

Output:

dist\BLACKBOX.exe

Privacy

Scanning happens locally. BLACKBOX does not require an account, backend, or source-code upload.

Scope

BLACKBOX is a first-pass forensic view, not a compiler-grade static analyzer. Findings are investigation signals rather than proof that code is unsafe or defective.

Release process

• Every push runs tests/compile checks and builds a Windows executable artifact.
• Tags matching v* build the executable again, compute SHA256, and publish both files to GitHub Releases.
• See CHANGELOG.md for release history.

License

MIT