Loading video...
Video Failed to Load
Here is how you can give Claude Code access to any data that exists online. It's an easy way to make it 10x more powerful than it already is. For example: Use Claude Code to find open LinkedIn jobs in your area, tailor your resume to them, and apply... show more
179,383 views • 9 months ago •via X (Twitter)
34 Comments

I'm using the @apify skills for Claude. Here is the GitHub repository: You can install these as a plugin, or you can also build your own skill to do whatever you want. Alternatively, you can install any of the scrapers using MCP directly in Claude Code.

You can browse the list of available actors on the platform: That's where you can create your own actors or expose existing ones through MCP. Thanks to the @apify team for collaborating with me on this post!

this is the workflow shift people underestimate - claude code with live data access turns "generate code" into "solve the actual problem." the context layer becomes more important than the model itself 🔗

You should be mentioning that Apify is not free and in many scenarios is quite expensive.

Great tutorial! Another complementary approach: Chrome MCP integration. We've been building multi-agent document intelligence where Claude Code orchestrates browser actions directly - navigate, screenshot, extract data from authenticated sessions (Supabase dashboards, X analytics, etc). Apify for scale scraping + MCP for interactive workflows = powerful combo.

it's because of shit like this that we have 10,000 applications on every single job posting five minutes after it's posted and nobody's even reading them. This is a dumb use case and you need to do better.

When using Claude Code to scrape online data, considering the website's terms of service is crucial to avoid getting blocked. You can also use it to find patterns in publicly available financial data, which can be particularly useful for market research. This approach can help you uncover valuable insights that might not be immediately apparent.

the skills approach is way more composable than just raw mcp. being able to install a scraper as a skill means you can chain it with other tools without writing glue code. curious if you've tried combining the apify actors with custom post processing skills to filter results before they hit your context window?

Interesting take! we built something similar to help pull the latest documents or web data into Claude code. It indeed makes it ten times more powerful and adds a lot more context.

have mutliple agents to it together in parallel with

Crazy

The automation arms race just turned job hunting into a spam factory. When everyone's resume gets perfectly tailored by AI, the signal disappears into the noise.

Nice

Your thread is creating a buzz! #TopUnroll 🙏🏼@Nidg31 for 🥇unroll

Its is looking interesting to watch

Man this is awesome! Thank you!

Sounds like a game changer for job hunting! Thanks for sharing!

the skills approach is way more composable than just raw mcp. being able to install a scraper as a skill means you can chain it with other tools without writing glue code. curious if you've tried combining the apify actors with custom post processing skills to filter results before they hit your context window?

Sneaky sneaky I like it

Thank for sharing @svpino We added 8 new skills in the last two days. Would love to hear your feedback which use cases and Actors we should add next 🙂

Agree-skills make this stuff really composable, you can chain scrapers with custom filters without writing a ton of glue code

mine

take it from someone who submits 5000 applications per week with a team of VAs - human in the loop and applying via career page is best - which I am sure is totally doable as part of this workflow.

Automating job searches sounds efficient but watch for data privacy.

Cost?

Bravvoooo sir thank you

wait claude code can actually scrape linkedin jobs and auto-apply now? tried it last week and it kept getting captcha'd to death lol

I hate modern job seeking. Fire up all your ai agents and spam all the job listings.

Giving models access to real-world data is the unlock. The real advantage comes from pairing that access with constraints, evaluation loops, and human judgment.

bro this is next level, gonna have Claude working harder than my coffee maker 😂

To save 5 mins to others - he doesn't show how to use Claude Code to find open LinkedIn jobs, doesn't show how to tailor your resume and doesn't show how to apply for them.

@clydedevv

this is an ad

tried feeding claude code a linkedin search url and it just hallucinated the whole job list lol

