Job hunting is exhausting, but not because applying for jobs is intellectually difficult. It drains you because the repetitive manual tasks, like constantly rewriting your resume bullets to match new descriptions, eat up hours of your time. The solution is Claude Cowork. It’s Anthropic’s desktop AI agent, and instead of acting like a chatbot in a browser window, it actually works directly alongside you on your local files. To get this running, you need three specific things: the Claude desktop app on a paid plan, a dedicated local folder on your computer, and a comprehensive master resume. Once set up, your role shifts entirely. You stop doing the manual formatting and data entry, and you step into the role of a manager who simply reviews and approves high-quality application packages. Mastering this workflow slashes your application time from 40 minutes per job down to under 10 minutes without sacrificing an ounce of quality.

Step one is setting up your environment. Create a dedicated job search folder on your computer, and inside it, add specific subfolders labeled Resumes, Cover Letters, and Applications. Next, you must explicitly grant the Claude Cowork agent permission to access and read this specific local directory. It cannot see these files until you allow it.

Now you write your prompt. You instruct Cowork to read a targeted job description in your folder and cross-reference its requirements against your master resume. You need to set strict constraints for the AI here: tell it to reorder your bullet points to match the role’s priorities, enforce a strict one-page limit, and explicitly tell it to never invent any experience. But the most critical instruction is demanding a changelog. You need a clear list of every edit the AI makes so you can verify its work. Within seconds, Cowork generates a brand new, correctly formatted resume file directly into your local subfolder alongside that detailed list of changes. This is where you have to watch out for AI hallucinations. You must review that changelog manually to ensure it hasn’t invented any metrics or exaggerated your accomplishments. Isolating the AI’s choices into a changelog guarantees you get an ATS-ready resume in seconds while keeping your professional history completely authentic.

Now let’s tackle cover letters. The problem with standard AI-generated cover letters is they read perfectly but say nothing. They are overly fluent, highly generic, and lack your specific human voice. You fix this using a voice reference strategy. Point Cowork to a subfolder containing your past cover letters along with a factual brag bank document full of your specific accomplishments. Set your constraints: limit the draft to 180 words, ban corporate clichés entirely, and instruct the AI to tie one specific achievement from your brag bank to a core requirement in the job posting. Because you force the AI to pull from your own files and match your tone, the output is an actual, specific letter that actually sounds like you wrote it.

With your materials prepped, you need to track them. It is incredibly easy to lose momentum when applications go quiet and get lost in the shuffle. Build an auto-updating tracker by telling Cowork to generate a CSV file in your folder with columns for company, role, status, and next step. Daily, tell Cowork which jobs you applied to and it updates automatically. If a row sits inactive for 7 days, it gets flagged. You then instruct Cowork to scan that tracker and automatically draft 7-day follow-up emails for any of those stalled applications. Building a system that updates its own records and drafts its own follow-ups prevents the administrative burnout that usually derails a job search.The final workflow prepares you for the interview itself. You can use Cowork’s sub-agent capability to run a dynamic 30-minute mock interview.

Ask Cowork to spin up distinct personas based on your resume and the job description: a manager, a tech lead, and a grader. The first sub-agent acts as the hiring manager; their job is to ask you behavioral and cultural fit questions. The second sub-agent plays a skeptical technical lead, pushing you hard on specific skill gaps and complex scenarios relevant to the role. The third sub-agent stays silent during the interview; it acts as a scorer, evaluating your answers strictly against the job description’s requirements. These agents ask questions sequentially, wait for your response, and then provide immediate critiques on how to tighten your answers before moving on. This multi-agent simulation forces you to articulate your thoughts out loud, transforming passive interview anxiety into active, highly targeted preparation.

We need to clear up the auto-apply myth right now. Cowork is an execution layer for your local files; it is not a discovery tool that will browse job boards and submit applications for you. By taking over the local file execution, this desktop engine handles all the manual formatting, data entry, and version control automatically. The hours you used to spend re-typing bullet points can now be redirected toward high-leverage activities like networking and identifying premium roles.

There’s a lot more to help you at jobsearch.community. Become an insider and get coaching from Jeff Altman, as well as video courses, books, and guides to help you, or you can purchase individual products and services à la carte. Jeff has helped tens of thousands of people land new jobs using techniques he learned while working in search. If you like the video, share it, leave a comment, click the like button, follow the channel—it will help other people if you do. If you want to thank me for the advice, buy me a coffee; the link is in the show notes.Tre ating the job search as an automated systems design problem is what ultimately separates exhausted candidates from those who secure top offers.

Why You Put Contact Info on Your Resume