How to Check Python with Ruff Online (ruff check, Tested)

ruff check is Ruff's linter: it parses your Python and flags issues — unused imports (F401), un-sorted imports (I001), deprecated patterns (UP…), likely bugs (B…), and hundreds more — each tagged with a rule code so you can look it up or silence it. It's the counterpart to ruff format (which only restyles code). Ruff is written in Rust, so a check that takes Flake8 seconds finishes in milliseconds.

1. Open the Ruff Playground — it loads the official Ruff WebAssembly build in your browser.
2. Paste or type your Python into the editor. Linting runs live as you type.
3. Read the Diagnostics tab: every issue shows its rule code, line:column, and a [*] marker if Ruff can auto-fix it.
4. Open Settings to edit the config as JSON — the same fields as [tool.ruff] in pyproject.toml (select, ignore, line-length).
5. Switch tabs to also see the Formatted output (ruff format), the AST, Tokens, or the Formatter IR.

That's the whole loop. Nothing is uploaded — the check runs on your machine — so you can safely paste proprietary code. Locally, the same command is ruff check path/to/file.py after pip install ruff (or uv add ruff).

Here's the file I used — intentionally sloppy, the kind of thing that accumulates in a real codebase:

import os, sys
import json
from typing import List

def load(path):
    f = open(path)
    data = json.load(f)
    names: List[str] = data['names']
    result = []
    for i in range(len(names)):
        result.append(names[i].upper())
    unused = 42
    return result

x=load('data.json')
print( x )

Running ruff check messy.py --select E,F,I,UP,B,SIM gives this — real, unedited output:

messy.py:1:1: E401 [*] Multiple imports on one line
messy.py:1:1: I001 [*] Import block is un-sorted or un-formatted
messy.py:1:8: F401 [*] `os` imported but unused
messy.py:1:12: F401 [*] `sys` imported but unused
messy.py:3:1: UP035 `typing.List` is deprecated, use `list` instead
messy.py:6:9: SIM115 Use a context manager for opening files
messy.py:8:12: UP006 [*] Use `list` instead of `List` for type annotation
messy.py:12:5: F841 Local variable `unused` is assigned to but never used
Found 8 errors.
[*] 5 fixable with the `--fix` option (1 hidden fix can be enabled with the `--unsafe-fixes` option).

Eight issues from sixteen lines. Notice the [*] markers — those five are auto-fixable. The playground shows the same list in its Diagnostics tab, with the [*] rendered as a fixable badge.

Add --fix and Ruff rewrites the safe issues in place. On my file:

Found 8 errors (6 fixed, 2 remaining).

Six gone in one pass. The two left (SIM115, the file opened without a context manager, and the hidden unsafe fix) need a human decision — Ruff won't guess there. The import block collapsed and sorted itself down to just what's used:

import json


def load(path):
    f = open(path)
    data = json.load(f)

ruff check finds problems; ruff format fixes *style*. They're separate commands (and separate tabs in the playground). Real before/after on a badly-spaced snippet:

def add(a,b) :
    return a+b
x = { "k":1,"y":2 }

def add(a, b):
    return a + b


x = {"k": 1, "y": 2}

Same output you'd get from Black on most files — see Ruff format vs Black for where they differ. In the playground, this is the Formatted tab.

Every diagnostic starts with a letter-prefixed code. The prefix tells you which family (and which flake8 plugin it replaces):

You control which families run with select/ignore in the config — E/F (pyflakes + pycodestyle) are the sane default; add I, UP, B, SIM for import sorting, upgrades, bug-bears and simplifications. Edit these live in the playground's Settings panel.

Checking Ruff online is perfect for a quick look — trying a rule, sharing a reproducible snippet, or checking code on a machine where you can't install anything. For day-to-day work you'll want it local (pip install ruff or uv add ruff) and wired into your editor and CI, so ruff check and ruff format --check run on every save and every push. The rules and output are identical; the online playground just skips the install.