[Bug]: Target_elements doesnt work

[jay377]

crawl4ai version

0.7.4

Expected Behavior

Hello,

I am having an issue using target_elements to only save certain content to markdown while still allowing it to view all links a do long crawls. It wont work in python scripts and i tested even the simplest possible CLI test and it doesnt work

crwl "https://ai.google.dev/gemini-api/docs" -c "target_elements=['.devsite-banner-announcement']" -o markdown --bypass-cache

That command provides the entire page content as output instead of the expected output below:

EXPECTED BEHAVIOR OUTPUT:

Gemini 2.5 Flash Image (aka Nano Banana) is now available in the Gemini API! Learn more

Current Behaviour

crwl "https://ai.google.dev/gemini-api/docs" -c "target_elements=['.devsite-banner-announcement']" -o markdown --bypass-cache

That command provides the entire page content as output still.

It should only be providing the following content:

Gemini 2.5 Flash Image (aka Nano Banana) is now available in the Gemini API! [Learn more]

I can confirm 100% the selector is correct because when i use css selector function it works fine:

(crawl4ai) jay@MSI:~/projects/crawl4ai$ crwl "https://ai.google.dev/gemini-api/docs" -c "css_selector=.devsite-banner-announcement" -o markdown --bypass-cache Gemini 2.5 Flash Image (aka Nano Banana) is now available in the Gemini API! Learn more

Is this reproducible?

Yes

Inputs Causing the Bug

crwl "https://ai.google.dev/gemini-api/docs" -c "target_elements=['.devsite-banner-announcement']" -o markdown --bypass-cache

Steps to Reproduce

crwl "https://ai.google.dev/gemini-api/docs" -c "target_elements=['.devsite-banner-announcement']" -o markdown --bypass-cache

Code snippets

crwl "https://ai.google.dev/gemini-api/docs" -c "target_elements=['.devsite-banner-announcement']" -o markdown --bypass-cache

OS

Windows 11 WSL 2

Python version

3.11

Browser

No response

Browser version

No response

Error logs & Screenshots (if applicable)

No response