链接到版本帮助页面确实提到了“最新版本”按钮，但这并没有给你一个下载链接。

https://github.com/reactiveui/ReactiveUI/releases/latest

为此，你需要首先获得最新的标签(如“GitHub URL For最新发布的下载文件?”所述):

latestTag=$(git describe --tags `git rev-list --tags --max-count=1`)

curl -L https://github.com/reactiveui/ReactiveUI/releases/download/$latestTag/ReactiveUI-$latestTag.zip

在PHP -重定向到最新版本下载。简单地放在你的网络空间

<?php

/**
 * Download latest release from github release articats
 * License: Public Domain
 */

define('REPO', 'imi-digital/iRobo');

$opts = [
    'http' => [
        'method' => 'GET',
        'header' => [
            'User-Agent: PHP'
        ]
    ]
];

$context = stream_context_create($opts);

$releases = file_get_contents('https://api.github.com/repos/' . REPO . '/releases', false, $context);
$releases = json_decode($releases);

$url = $releases[0]->assets[0]->browser_download_url;

header('Location: ' . $url);

只需使用下面的网址下载最新版本: (以boxbilling项目中的url为例):https://api.github.com/repos/boxbilling/boxbilling/releases

以zip格式下载最新版本: https://api.github.com/repos/boxbilling/boxbilling/zipball

下载最新版本为tarball: https://api.github.com/repos/boxbilling/boxbilling/tarball

点击其中一个网址立即下载最新版本。当我写这一行时，它目前是:boxbilling-boxbilling-4.20-30-g452ad1c[.zip/.tar.gz]

更新:在我的日志文件中发现了另一个url(参考上面的例子) https://codeload.github.com/boxbilling/boxbilling/legacy.tar.gz/master

链接到版本帮助页面确实提到了“最新版本”按钮，但这并没有给你一个下载链接。

https://github.com/reactiveui/ReactiveUI/releases/latest

为此，你需要首先获得最新的标签(如“GitHub URL For最新发布的下载文件?”所述):

latestTag=$(git describe --tags `git rev-list --tags --max-count=1`)

curl -L https://github.com/reactiveui/ReactiveUI/releases/download/$latestTag/ReactiveUI-$latestTag.zip

这是针对Linux的。

我看到了上面被接受的答案

A few years late, but I just implemented a simple redirect to support https://github.com/USER/PROJECT/releases/latest/download/package.zip. That should redirected to the latest tagged package.zip release asset. Hope it's handy!

但有评论指出，它不支持版本化文件名。

在搜索了一会儿之后，我编写了一个适用于版本化文件名的一行调用。它使用curl获取最新的文件版本，然后使用添加的重定向支持来下载最新版本的文件。

wget $'https://github.com/<UMBRELLA PROJECT>/<REPO NAME>/releases/latest/download/<FILE NAME START>-'$(curl -s https://github.com/<UMBRELLA PROJECT>/<REPO NAME>/releases/latest | grep -o -P '(?<=releases/tag/).*(?=\">)')$'<FILE NAME END>'

So it targets a file that's named like <REPO NAME>-linux64_arm-<VERSION NUMBER>.tar.gz that's on the webpage https://github.com/<UMBRELLA PROJECT>/<REPO NAME>/releases/latest after it redirects. It does this by looking for the <VERSION NUMBER> between releases/tag/ and the "> in the text that's returned from the curl call. So to be really explicit, <FILE NAME START> is the <REPO NAME>-linux64_arm- and <FILE NAME END> is the .tar.gz in the above example. Get the START and END bits by looking at what the https://github.com/<UMBRELLA PROJECT>/<REPO NAME>/releases/latest uses as its file naming scheme.

我通过模仿其他人如何使用grep和curl来创建这个，现在才学会所有这些，所以如果它做了一些我甚至无法理解的真正淘气的事情，请告诉我!我还说<UMBRELLA PROJECT>，但用户名应该能够去那里就好了。大声喊出https://stackoverflow.com/a/13245961/2403531的grep调用，https://unix.stackexchange.com/a/10264的$字符串$连接。

使用(内部)wget获取HTML内容，过滤zip文件(使用egrep)，然后下载zip文件(使用外部wget)的解决方案。

wget https://github.com/$(wget https://github.com/<USER>/<PROJECT>/releases/latest -O - | egrep '/.*/.*/.*zip' -o)