How to Use Claude to Screen Resumes: The Alex Skill Guide

Fifty applications, five interview slots, and an afternoon to bridge the gap. This is where a lot of hiring quietly goes wrong — not in the interview, but in the rushed, inconsistent first cut. AI resume screening promises to fix it, and a raw chatbot half-does: paste a CV, ask “is this any good?”, and you get a confident answer with no criteria behind it. Ask it about ten candidates the same way and you get ten answers judged by ten subtly different standards.

Alex is the fix. It is a resume-screening persona you load once into Claude, ChatGPT, or any AI chat, and its first move is the one most people skip: it refuses to judge a CV until it knows what it is judging against. Define the role and the must-haves first, and every candidate after that is measured by the same ruler. Screening without criteria is just bias with a spreadsheet; this is the opposite.

Why generic AI screening is risky

A bare chatbot screens inconsistently because nothing anchors it. The standard drifts from candidate to candidate, so the tenth CV is not judged like the first. Worse, an unguided model can quietly reproduce bias — reacting to a name, a university, or an employment gap that has nothing to do with whether someone can do the job. And it tends to be over-confident, handing down a clean “reject” where a careful human would have flagged something worth a five-minute phone call.

What changes with the Alex skill

Criteria come first. Alex asks for the role brief and separates the genuine must-haves from the nice-to-haves before it reads a single application. It then scores each candidate against that rubric and returns a structured verdict — advance, hold, or no — with two or three concrete reasons, not an essay. Red flags are framed as questions to explore (“three short tenures in a row — worth asking about in a screen call”) rather than automatic disqualifiers. And it evaluates against job-relevant evidence only, leaving names, schools, and career gaps out of the judgement unless they bear directly on the role.

What it actually produces

You get a screening rubric you can reuse across a whole role, a ranked shortlist with an advance/hold/no call and the reasoning for each, and a green-flag and red-flag breakdown on the candidates worth a closer look. The output is built to be scanned — candidate, verdict, top reasons — so a stack of fifty becomes a shortlist of five in one pass rather than one long afternoon.

How to get the most out of it

Spend the first two minutes on criteria, not CVs. Be explicit about what is non-negotiable and what is merely desirable, and give it the actual job brief rather than a job title. Let it surface concerns as questions, and keep the final decision with a person — the skill sharpens and speeds up the first cut, but a hire is yours to make. Used this way it removes the drudgery without removing the judgement.

Who this is for

In-house recruiters working through high application volumes, founders making their first hires without a talent team, and hiring managers who want a faster, fairer first screen. It works with Claude, ChatGPT, or any AI chat that accepts a system prompt. If you want to cover the rest of the hiring process too, the work & business skills collection includes interview preparation, job descriptions, and onboarding — each a focused assistant rather than a general chatbot.