Installer vulture avec Homebrew, apt, MacPorts, pacman

Liens du projet

• Page d'accueil
• Docs amont
• Page du gestionnaire de paquets

Commandes et alias

• vulture

Historique du projet

Vulture has a longer history than its current GitHub repository suggests: PyPI records version 0.1 uploaded on March 17, 2012. The current GitHub repository was created in March 2017 under maintainer Jendrik Seipp, and the package metadata continues to identify him as the author/owner.

The project matured from a small dead-code finder into a production/stable Python quality tool. Its README emphasizes that Python's dynamic nature makes perfect static dead-code detection impossible, so Vulture reports confidence values and provides practical suppression mechanisms rather than pretending to be exact.

The 2.x series made Vulture fit modern Python project workflows: it supports `pyproject.toml` configuration, pre-commit integration, programmatic use, whitelists for false positives, common whitelists in the repository, and output syntax that complements Pyflakes.

Historique d'adoption

Vulture became a recognizable tool in Python maintenance circles because it addresses a recurring problem in large, long-lived codebases: unused helpers, stale imports, abandoned feature paths, and untested code that ordinary linting or test coverage may not make obvious.

By July 2026, GitHub repository metadata showed more than 4.6k stars, and PyPI listed Vulture 2.16 as the current release, uploaded on March 25, 2026. It also appears in editor and automation workflows through pre-commit hooks, a GitHub Action, and VS Code extension integrations.

Modes d'utilisation

Typical usage is `vulture path_or_file`, optionally with directories, a generated whitelist, and a confidence threshold such as `--min-confidence 100`. The tool recommends scanning both a library and its test suite so code used only by tests is seen as used, and then rerunning after deletions because removing one chunk can reveal more dead code.

Vulture's confidence model is part of its identity. It treats unreachable code and unused function or method arguments as 100 percent, imports as 90 percent, and many attributes, classes, functions, methods, properties, and variables as 60 percent. Users tune results with whitelists, excludes, ignored names or decorators, `noqa` compatibility for selected Pyflakes codes, and `pyproject.toml` settings.

Pourquoi les passionnés de paquets s'y intéressent

For package nerds, Vulture is the small sharp tool between linting and coverage. It is simpler than a whole quality platform, depends mostly on Python's parser, and is valuable in cleanup branches, CI gates, and pre-commit checks where the goal is to find deletion candidates before they fossilize.

Chronologie

• 2012: Vulture 0.1 was uploaded to PyPI on March 17, 2012.
• 2017: The current GitHub repository was created on March 6, 2017.
• 2021: Vulture 2.3 appeared as the example pre-commit hook revision in the README's integration documentation.
• 2026: Vulture 2.16 was uploaded to PyPI on March 25, 2026.

Related projects

• Vulture is related to Pyflakes, Flake8, Coverage.py, pre-commit, Ruff discussions around dead-code detection, editor extensions that wrap Vulture, and other Python dead-code tools such as uncalled and dead.

Sources

• https://api.github.com/repos/jendrikseipp/vulture
• https://github.com/astral-sh/ruff/issues/872
• https://github.com/jendrikseipp/vulture
• https://pypi.org/project/vulture/
• https://pypi.org/pypi/vulture/json
• https://raw.githubusercontent.com/jendrikseipp/vulture/main/README.md