Right-click any file in Windows Explorer, pick the format you want, and get a converted copy instantly.
$curl -sLO https://github.com/Ameya79/cuttle/releases/latest/download/Cuttle-Setup-x64.exe
Where useful open source lives.
Small.Onejob.Workswhenyouinstallit.
Every project gets the same checks.
Automated tests verify licenses, package builds, active releases, and clean installation into fresh environments.
rollitAutomated run
[pass]README
Present[pass]License
MIT[pass]Tagged release
v1.0.1[pass]Last commit
128d ago[pass]Installs clean
pip install & import ok[pass]Tests pass
GreenWheel size: 8.1 KBDependencies: 1
Result6 of 6 passed
Sample runUnresolved dependency
[pass]README
Present[pass]License
MIT[pass]Tagged release
v0.9.0[fail]Installs clean
ModuleNotFoundErrorExample. A check that can't fail isn't a check.
The Benchtop Thesis
An automated verification authority for focused open source software.
Modern software development moves fast, but package reliability has lagged behind. We evaluate tools against reproducible, automated criteria so developers never have to guess whether an installation will succeed.
01 / The Registry Dilemma
Cutting Through Hype and Bloat
Public package registries are saturated with abandoned repositories, hidden breaking changes, and deep dependency chains. Vanity metrics like GitHub stars and social buzz reward marketing over dependable engineering. We focus strictly on verifiable execution health.
02 / Automated Isolation
Reproducible Sandbox Testing
Every evaluated project runs through ephemeral container environments with zero network leakage during install verification. We verify that packages install cleanly into fresh runtimes, enforce permissive licenses, maintain active commits, and pass their own test suites.
03 / The Curated Commons
Promoting Independent Craft
Benchtop Labs exists to elevate single-purpose utilities and focused libraries built by independent creators. By establishing transparent benchmarks, we provide high-craft open-source projects the visibility and trust they deserve without requiring corporate sponsorship.
Curated Open Source Index
On the bench
Rolling window statistics for numpy arrays.
$pip install rollit
Automatic spending limits on OpenAI, Claude, and 60+ AI models.
$pip install "talanton-py[openai]"
Non-AI static analysis layer for AI coding agents with token-efficient diagnostics.
$pip install mustel
Stateless Monte Carlo stock price simulation and quantitative risk engine API in Python.
$git clone https://github.com/Ameya79/Driftwood.git
Submissions Open
Put your project on the bench
Submit your open source project for automated verification. Every project runs through our transparent 6-point bench test in an isolated sandbox. Zero failing checks gets you listed in the directory and certified with a verified README badge.