agentleFS
Sign inStart free
OneWave-AIOneWave-AI/claude-skills/cowork-hiring-screener/SKILL.md

cowork-hiring-screener

Skill378 starsChanged 54 days agoMITView on GitHub

Point Cowork at a folder of resumes plus a job description -- screens every candidate against the actual requirements, produces a ranked shortlist with evidence, drafts advance/decline emails, and builds interview kits for the top picks. Pairs with hiring-scorecard for the interview stage.

The file

We split this skill into the rules it gives an agent

3 rules. About 0 are common, 3 are distinctive, and 0 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈3

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 0

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Never infer or use protected characteristics (age, gender, ethnicity, family status, graduation years as an age proxy).

    line 24
  2. Distinctive:

    Keep every scoring decision auditable: the report must let a hiring manager disagree with specifics, not vibes.

    line 27
  3. Distinctive:

    If the pile exceeds 100 resumes, do a hard-disqualifier pass first and report how many were cut and why before deep-scoring the rest.

    line 28

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Cowork Hiring Screener

Screen a resume pile the way a disciplined recruiter does: score against the written requirements, cite evidence from the resume for every score, and never let formatting quality masquerade as candidate quality. Input is a folder of resumes (PDF, .docx, text) and a job description; output is a defensible shortlist.

Workflow

  1. Extract requirements. Parse the JD into must-haves, nice-to-haves, and disqualifiers. Present the rubric for approval before scoring -- the human may reweight. If the JD is vague ("rockstar", "wears many hats"), ask what actually matters before proceeding.
  2. Inventory. Catalog every file in the folder. Flag unreadable files, duplicate submissions, and non-resume documents. Report the candidate count before starting.
  3. Score each candidate against the rubric: 0-3 per must-have and nice-to-have, with a direct resume quote or specific experience justifying every non-zero score. No quote, no points.
  4. Rank and tier. Produce screening-report.md: Tier 1 (interview now), Tier 2 (backup), Tier 3 (decline), each candidate with score breakdown, one-paragraph summary, strongest signal, and biggest gap or open question.
  5. Draft communications. Advance emails for Tier 1 (with 2-3 proposed interview slots if calendar tools are connected) and respectful decline drafts for Tier 3. Drafts only -- never send.
  6. Interview kits. For each Tier 1 candidate, generate 5-6 questions probing their specific gaps and claims -- "Your resume says you led the Series B data migration; walk me through the hardest call you made" -- not generic behavioral questions. Hand off to hiring-scorecard for structured interview evaluation.

Rules

  • Score the content, not the polish. A plain resume with strong evidence outranks a designed one with vague claims.
  • Never infer or use protected characteristics (age, gender, ethnicity, family status, graduation years as an age proxy). Score skills and experience only.
  • Distinguish "did the thing" from "was near the thing." "Led migration" and "team migrated during my tenure" are different scores.
  • Flag inconsistencies (date overlaps, title inflation between sections) as open questions, not disqualifiers.
  • Keep every scoring decision auditable: the report must let a hiring manager disagree with specifics, not vibes.
  • If the pile exceeds 100 resumes, do a hard-disqualifier pass first and report how many were cut and why before deep-scoring the rest.

Quick Commands

  • "Screen [folder] against [JD]" -- full workflow
  • "Just the rubric" -- step 1 only, for approval
  • "Top 5 only" -- deep-score and report only the strongest candidates
  • "Draft the declines" -- Tier 3 communication drafts

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

Check your setup, free