"Cannot read property of undefined" when using chrome.tabs or other chrome API in content script

chrome.tabs returns undefined despite the fact I set tabs in the permissions block.

"permissions": [
    "tabs",
    "http://*/*",
    "https://*/*"
],
"content_scripts": [
    {
        "matches": [
            "http://*/*",
            "https://*/*"
        ],
        "js": [
            "js/myScript.js"
        ],
        "all_frames": true
    }
],

But in myScript.js the following returns undefined.

chrome.tabs   

As content script has its own limitations,

chrome.tabs is only available in background scripts and popup scripts.

If you wanna to use chrome.tabs then pass message from content_script to background script and play with chrome.tabs.

Content scripts have only limited access to Chrome APIs. This access does not include the API you are trying to use (e.g. chrome.tabs). If you need to use that API, you will have to do so in a background script¹.

As listed in Chrome's content scripts documentation, the APIs available to a content script are [I have placed deprecated methods in strikethrough format]:

• extension ( getURL , inIncognitoContext , lastError , onRequest , sendRequest )
• i18n
• runtime ( connect , getManifest , getURL , id , onConnect , onMessage , sendMessage )
• storage

A couple of the listed APIs are deprecated and have been for some time. Those that are deprecated have moved to different locations (also listed above):

• extension.onRequest ➞ runtime.onMessage
• extension.sendRequest ➞ runtime.sendMessage

While not officially deprecated, extension.lastError is also available as runtime.lastError. At this point, it is usually referred to at that location:

• extension.lastError ➞ runtime.lastError

Partition your extension into background scripts and content scripts

You are going to need to separate your code into what needs to be in a background script and what needs to be in content scripts, based on the capabilities available to each type of script. Content scripts have access to the DOM of the web page into which they are injected, but limited access to extension APIs. Background scripts have full access to the extension APIs, but no access to web page content. You should read the Chrome extension overview, and the pages linked from there, to get a feel for what functionality should be located in which type of script.

It is common to need to communicate between your content scripts and background scripts. To do so you can use message passing. This allows you to communicate information between the two scripts to accomplish things which are not possible using only one type of script. For instance, in your content script, you may need information which is only available from one of the other Chrome APIs, or you need something to happen which can most appropriately (or only) be done through one of the other Chrome extension APIs. In these cases, you will need to send a message to your background script, using chrome.runtime.sendMessage(), to tell it what needs to be done, while providing enough informaiton for it to be able to do so. Your background script can then return the desired information, if any, to your content script. Alternately, you will have times when the processing will primarily be done in the background script. The background script may inject a content script, or just message an already injected script, to obtain information from a page, or make changes to the web page.

1. Background script means any script that is in the background context. In addition to actual background scripts, this includes popups and options pages, etc. However, the only page that you can be sure to have consistently available to receive messages from a content script are your actual background scripts defined in manifest.json. Other pages may be available at some times as a result of the user's interaction with the browser, but they are not available consistently.

^{This answer was moved from a duplicate question, and then modified.}