Recruiters do not need another blank chat box. They need a way to make the assistant behave like it understands the hiring workflow before the conversation starts.

That is the point of Ribbon's recruiting skills directory. It is a field directory of reusable recruiter skills that can be copied into ChatGPT, Claude, or any assistant that accepts custom instructions. The page currently organizes 116 skills across 12 recruiting lanes, from planning and sourcing to screening, interviewing, offers, analytics, governance, and role-specific hiring.

A skill is not a magic prompt. It is a small operating procedure. It tells the assistant when to use the skill, what inputs to collect, what workflow to follow, what output to return, and what guardrails should stay in place. That matters because recruiting work is full of repeated judgment calls. The assistant should not guess the hiring plan from scratch every time.

Why recruiter skills matter

Most teams start with generic prompts: write an outreach message, screen this resume, summarize this interview, build a Boolean search. Those prompts work once, then drift. One recruiter asks for a short list. Another asks for a market map. A third forgets to include location, compensation, must-have criteria, or stop rules. The assistant gives each person a plausible answer, but the workflow is different every time.

Skills make the repeatable part explicit. They give the assistant a shared pattern for the task. For recruiting teams, that means less time re-explaining the basics and more time reviewing the actual judgment: Is this role defined clearly? Are the must-haves job related? Are we over-weighting a weak signal? Did we preserve a human decision point before a rejection or offer?

This is especially useful in high-volume hiring and agency work, where the same motion repeats across many roles. A sourcing skill can keep title variants, target companies, and exclusion logic in one place. A screening skill can require evidence and assumptions to be separated. A governance skill can remind the assistant not to infer protected characteristics or invent candidate facts.

What the Ribbon skills page includes

The directory is built for recruiters who want useful instructions without writing them from a blank page. Each card has a plain-English description, a category, tags, and source inspiration. When a recruiter expands a card, Ribbon turns it into a complete instruction set that can be copied into an assistant.

The skills cover the real lanes of recruiting work:

• Planning: intake calibration, scorecards, candidate personas, and job description cleanup.
• Sourcing: market mapping, Boolean strings, LinkedIn, GitHub, open web sourcing, and community search.
• Screening: evidence extraction, fit summaries, risk flags, and follow-up questions.
• Engagement: outreach, nurture, reactivation, scheduling, and candidate communication.
• Interviewing and offer: structured interview topics, debrief preparation, offer-risk planning, and close plans.
• Analytics and governance: funnel diagnosis, quality checks, audit trails, privacy, bias review, and human approvals.

The value is not just the list. It is the shape. Every skill returns a practical artifact first, then evidence and assumptions, then risks, missing context, and human-review points. That order helps a recruiter skim fast without losing the parts that should be checked carefully.

How to use a skill in ChatGPT or Claude

The simplest path is to open the skill, copy the instruction set, and paste it into a new chat before adding the role or candidate context. That works for one-off tasks.

For repeat work, put the skill somewhere more durable. In ChatGPT, that might be a Project instruction, a custom GPT instruction, or a saved custom instruction depending on how broad you want the behavior to be. In Claude, it might be Project instructions for a hiring workspace, a reusable Claude Code skill if your team works from files, or a normal chat when you only need the workflow once.

Scope matters. Do not put every recruiting skill into a global instruction field. A global instruction should be stable and broad, such as tone or preferred output format. A sourcing skill should live near sourcing work. A screening skill should live near role scorecards, resumes, transcripts, or interview notes. The more specific the skill, the more specific the context should be.

There is also a useful distinction between skills and connected data. Ribbon's MCP page lets a recruiter work with read-only ATS data from Claude and other clients that support remote MCP. A skill tells the assistant how to think through the work. MCP gives the assistant permissioned data to read. Used together, they make the assistant more practical: it can follow a recruiting workflow and work from the right system context.

Example: use the Market mapping skill

Imagine you need to source a hard-to-fill maintenance supervisor role across three cities. A generic prompt might produce a list of obvious companies. The Market mapping skill pushes the assistant to do more structured work.

Skill to use: Market mapping.

When to use it: map target companies, adjacent industries, role title variants, and talent density for a hard-to-fill role.

Inputs to collect:

• Role title, location, seniority, must-have skills, nice-to-have skills, and disqualifiers.
• Target industries, adjacent industries, known competitors, and companies to avoid.
• Pay range, shift requirements, travel requirements, work authorization constraints, and urgency.
• Any existing pipeline data, response-rate notes, or hiring manager preferences.

Workflow to paste into the assistant:

1. Restate the role and ask for missing inputs before building the map.
2. Generate title variants and adjacent titles that could represent the same work.
3. Group target companies into direct competitors, adjacent employers, training grounds, and less obvious sources.
4. For each group, explain why candidates from that lane may or may not fit.
5. Return search strings, example outreach angles, and a recruiter review checklist.

Example prompt:

Use the Market mapping skill. We are hiring a Maintenance Supervisor for multi-site facilities in Dallas, Fort Worth, and Arlington. Must-haves: preventive maintenance, vendor coordination, basic electrical and HVAC familiarity, and experience covering more than one location. Avoid candidates whose only experience is residential handyman work. Build a target company map, title variants, Boolean searches, and a review checklist for the recruiter.

A good answer should not simply hand back names. It should show the sourcing logic. It should separate direct competitors from adjacent talent pools, flag assumptions about location and seniority, and make it easy for the recruiter to reject weak lanes before spending time on outreach.

Where skills fit in a recruiting workflow

Skills are best used at moments where the assistant needs structure more than creativity. Intake, sourcing, screening, interview preparation, candidate summaries, and funnel diagnosis all benefit from a repeatable path. The recruiter still owns the decision. The skill just reduces the amount of setup required to get a useful first pass.

They also help teams train their own habits. A good skill reminds the recruiter to collect the right inputs before asking for output. It forces a distinction between facts and assumptions. It asks for risks and missing context. It keeps human review visible. Those are not small details. They are what separate an assistant that writes plausible text from an assistant that supports a hiring operation.

The next version of recruiting work will not be one giant prompt that does everything. It will be a set of small, portable procedures that teams can reuse, inspect, and improve. That is why the skills page is useful: it gives recruiters a practical starting library, then lets them bring those workflows into the assistants they already use.