Claude Skills for Recruiters: Where They Help and Where You Stay in Charge

By F. Jay Hall · Published November 6, 2026

I have recruited executives for 28 years, and I build with AI every day to run ExecSearches. So when HeroHunt published its list of the top ten Claude skills for recruiting, I read it closely. It is a useful list of prompts. It also blurs two different things, and the difference matters if you plan to put this to work in a search.

A prompt is something you type. A Claude Skill is something you build once and reuse. This article explains what a Skill actually is, the recruiting work where one earns its keep, how to build a simple one, and the places where a person has to make the call.

What a Claude Skill actually is

Anthropic introduced Agent Skills in October 2025. A Skill is a folder. Inside it sits a file called SKILL.md, which opens with a short header giving the Skill a name and a description, followed by plain instructions. The folder can also hold templates and small scripts.

The useful part is how Claude loads them. At the start, Claude only sees each Skill’s name and description. When your request matches a description, it reads the full instructions, and it opens the extra files only if the task needs them. Anthropic calls this progressive disclosure. In practice it means you can keep a shelf of Skills installed without cluttering every conversation.

Anthropic ships ready-made Skills for Word, Excel, PowerPoint and PDF files. You can also write your own. Custom Skills run in the Claude apps (on paid plans with code execution turned on), in Claude Code, and through the Claude API. They do not sync between those places, so a Skill you upload to the Claude app has to be uploaded separately to the API. In December 2025 Anthropic released the format as an open standard, and other agent tools now read it too.

One more thing from Anthropic’s own documentation deserves attention: install Skills only from sources you trust. A Skill can carry code, and a bad one can direct Claude to do things you never intended. Read every file in a Skill before you use it, the same way you would read a contract.

Where a Skill earns its keep in recruiting

The best candidates for a Skill are tasks you do the same way every time, where your standard matters more than the wording of any single request. In executive and nonprofit search, these are the ones I would start with.

Job descriptions. Write your house format into the Skill once: the order of sections, how you state scope and the reporting line, and the words you never use. Every draft comes out in your structure. Our nonprofit executive job description builder follows the same idea in a simpler form.

Intake notes. After a kickoff call with a board chair or hiring manager, a Skill can turn your rough notes into a clean position profile with the open questions listed at the end. The open questions are the valuable part.

Candidate research summaries. A Skill can pull public information you provide into a one-page brief in the same order every time: career arc, scope of past roles, board experience, published work. You still decide what is relevant and what is nobody’s business.

Outreach drafts. Give the Skill your voice rules and a few examples of messages that worked. It drafts, and you edit and send, so nothing goes out on its own.

Interview guides and scorecards. Put the competencies for the role and your rating scale in the Skill, and it produces a structured guide with follow-up questions for each competency. Structured interviews are fairer to candidates, and a Skill makes them easy to run consistently.

Reference-check summaries. A Skill can organize what references actually said under each competency, quoting rather than characterizing. That makes it harder for a summary to drift from the source.

Compliance checks on your own documents. A Skill can review a job post against a checklist you maintain: is the salary range stated where the law requires it, is the language plain, are the requirements real requirements. It flags problems and leaves the decision to you.

Notice what is missing from that list: ranking candidates and deciding who moves forward. I will come back to that.

How to build a simple one

You do not need to write code to start. In the Claude app, Anthropic provides a skill-creator Skill that interviews you about your workflow, builds the folder, and formats the SKILL.md file. On Team and Enterprise plans, an administrator has to turn Skills on for the organization first.

If you prefer to write it yourself, keep it plain. The header needs a name of up to 64 characters, using lowercase letters, numbers and hyphens, and a description of up to 1,024 characters that says what the Skill does and when to use it. A first recruiting Skill might look like this in outline:

name: nonprofit-job-description

description: Drafts a nonprofit executive job description in our house format. Use when asked to write or revise a job posting for an executive, director or C-suite role.

Then the instructions: the sections in order, the facts to ask for when they are missing, plus one finished example. Anthropic’s guidance is to keep the main file concise and under 500 lines, and to move long reference material into separate files that Claude opens only when it needs them.

Test it on three real searches before you rely on it. When it gets something wrong, fix the instructions rather than the output, so the next draft comes out right.

Where a person has to stay in charge

At ExecSearches, tools propose and a person approves. That rule is not caution for its own sake. In recruiting, it is where the law is heading.

New York City’s Local Law 144 already requires employers and employment agencies that use an automated employment decision tool to have it bias-audited within the past year, publish information about that audit, and notify candidates. The EU AI Act lists AI used for recruitment and selection, including filtering applications and evaluating candidates, among its high-risk uses. If a Skill screens or scores people and that score shapes who advances, you may be operating exactly the kind of tool these rules cover.

So keep a person on these decisions:

Who advances. A Skill can organize what you know about a candidate. It should not rank people and move them forward or out without a recruiter reading the file.

What you keep. Candidate information belongs in the systems you control. Do not paste a candidate’s confidential details into tools or Skills you have not vetted, and remove them when the search closes.

What you tell people. If AI helped draft a message or summary that a candidate or client sees, be ready to say so. I would rather explain how we work than have someone find out later.

What the client hears. The final recommendation to a board is a judgment about people and fit. That part is the job.

Where this leaves nonprofit recruiters

For a small search firm or a nonprofit HR team of one, Skills are a real gain. They turn your best habits into something consistent that a new colleague, or you on a long day, can use the same way every time. Start with the job description or the intake note, keep a person on every decision about a human being, and read every Skill before you install it.

If you want the broader picture of building with AI, my eight-step field guide to building an AI agent covers the next step up. And if your job posts are not landing, the same discipline applies to the writing, and Copywriting Tips for Nonprofits covers it.

Frequently asked questions

What is the difference between a Claude Skill and a prompt?

A prompt is a one-time instruction you type. A Skill is a saved folder of instructions, and sometimes templates or scripts, that Claude loads automatically whenever a request matches its description.

Do I need to code to build a Claude Skill?

No. In the Claude app, Anthropic’s skill-creator Skill can build one with you through conversation. Writing a SKILL.md file by hand only takes a plain-text editor.

Can a Claude Skill screen or rank job candidates?

It can be built to, but I advise against letting it decide who advances. Laws such as New York City’s Local Law 144 and the EU AI Act place specific obligations on automated tools used in hiring decisions. Use Skills to organize information and draft documents, and keep a recruiter on every decision about a candidate.

Are Claude Skills safe to install?

Anthropic recommends installing Skills only from trusted sources, such as ones you wrote or ones from Anthropic, and reading every file in a Skill before using it, because a Skill can include code.

Sources

• HeroHunt, “Top 10 Claude Skills for Recruiting (2026)”
• Anthropic, “Introducing Agent Skills” (October 16, 2025, with later updates)
• Anthropic Engineering, “Equipping agents for the real world with Agent Skills”
• Claude Platform Docs, “Agent Skills overview”
• Claude Platform Docs, “Skill authoring best practices”
• Claude Help Center, “What are skills?”
• Agent Skills open standard
• NYC Department of Consumer and Worker Protection, “Automated Employment Decision Tools”
• EU AI Act, Annex III: High-Risk AI Systems