@arkone_ai/cv-screen 1.0.0

package/README.md

@@ -0,0 +1,28 @@
+# @arkone_ai/cv-screen
+
+A [Claude Code](https://claude.ai/code) skill that screens a candidate's CV/resume against a job description and produces a structured fit analysis.
+
+## Install
+
+```bash
+npm install -g @arkone_ai/cv-screen
+```
+
+## Usage
+
+In any Claude Code session:
+
+```
+/cv-screen
+```
+
+## What it does
+
+- Reads a CV (PDF, DOCX, image, or text) and a job description
+- Scores the candidate across 5 categories: skills match, experience, education, career trajectory, and role alignment
+- Detects red flags (job hopping, gaps, overqualification) neutrally as interview discussion points
+- Produces a structured report with overall score, recommendation (advance/hold/pass), and suggested interview focus areas
+
+## License
+
+MIT

package/install/install.js

@@ -0,0 +1,38 @@
+#!/usr/bin/env node
+
+import { copyFileSync, mkdirSync, existsSync } from 'fs';
+import { join } from 'path';
+import { homedir } from 'os';
+
+const root = join(import.meta.dirname, '..');
+const home = homedir();
+
+const agents = [
+  { name: 'Claude Code', dir: join(home, '.claude', 'skills') },
+  { name: 'Codex',       dir: join(home, '.codex',  'skills') },
+  { name: 'Kiro',        dir: join(home, '.kiro',   'skills') },
+];
+
+const files = [
+  { src: join(root, 'skill.md'), dest: 'cv-screen.md' },
+];
+
+let installed = 0;
+for (const agent of agents) {
+  if (!existsSync(agent.dir)) continue;
+  for (const file of files) {
+    copyFileSync(file.src, join(agent.dir, file.dest));
+  }
+  console.log(`✓ ${agent.name}: ${agent.dir}`);
+  installed++;
+}
+
+if (installed === 0) {
+  mkdirSync(agents[0].dir, { recursive: true });
+  for (const file of files) {
+    copyFileSync(file.src, join(agents[0].dir, file.dest));
+  }
+  console.log(`✓ Claude Code: ${agents[0].dir}`);
+}
+
+console.log('\nUse /cv-screen in any supported AI session.');

package/package.json

@@ -0,0 +1,27 @@
+{
+  "name": "@arkone_ai/cv-screen",
+  "version": "1.0.0",
+  "description": "Claude Code skill: screen a candidate CV against a job description with structured fit analysis",
+  "type": "module",
+  "scripts": {
+    "postinstall": "node install/install.js"
+  },
+  "bin": {
+    "cv-screen": "install/install.js"
+  },
+  "files": [
+    "skill.md",
+    "install/"
+  ],
+  "keywords": [
+    "claude",
+    "claude-code",
+    "arkone",
+    "hr-ops",
+    "cv-screen"
+  ],
+  "license": "MIT",
+  "engines": {
+    "node": ">=18"
+  }
+}

package/skill.md

@@ -0,0 +1,184 @@
+---
+name: cv-screen
+description: "Screen a single candidate CV/resume against a job description and produce a structured fit analysis. Use this skill when: 'screen this CV', 'review this resume', 'check candidate fit', 'analyze this CV against JD', 'evaluate this candidate'. Also trigger for: 'resume review', 'candidate assessment', 'CV analysis'."
+license: MIT
+compatibility: Designed for Claude Code
+---
+
+# Skill: cv-screen
+
+Screen a single candidate's CV/resume against a job description and produce a detailed, structured fit analysis.
+
+## When to use
+
+- User provides a CV and a JD and wants to know if the candidate is a fit
+- User asks to review or evaluate a resume
+- User wants interview focus areas based on a CV
+
+## Step 1: Gather inputs
+
+### Required
+
+| Input | Accepts |
+|---|---|
+| CV / Resume | File path (PDF, DOCX, TXT, image) or pasted text |
+| Job Description | File path (PDF, DOCX, TXT) or pasted text or URL |
+
+### Optional
+
+| Input | Default | Purpose |
+|---|---|---|
+| Must-have skills | Extracted from JD | Override or supplement skills parsed from JD |
+| Experience minimum | Extracted from JD | Override min years of experience |
+| Priority weights | Equal weight | Weight categories differently (e.g., skills 40%, experience 30%, education 20%, culture 10%) |
+| Screening strictness | Standard | `strict` = hard cutoffs, `standard` = balanced, `lenient` = focus on potential |
+
+### Reading the documents
+
+- **PDF**: Read directly using the Read tool (Claude can read PDFs natively)
+- **DOCX**: Use `python3 -c "import docx; ..."` or `textutil -convert txt <file>` (macOS) to extract text
+- **Image** (photo of CV): Read directly using the Read tool (Claude can read images natively)
+- **Pasted text**: Use as-is
+
+## Step 2: Parse the job description
+
+Extract and structure these elements from the JD:
+
+| Element | What to extract |
+|---|---|
+| Role title | Exact title |
+| Level / seniority | Junior, Mid, Senior, Lead, Manager, Director |
+| Must-have skills | Hard requirements (languages, frameworks, certifications) |
+| Nice-to-have skills | Preferred but not required |
+| Experience requirement | Minimum and preferred years |
+| Education requirement | Degree level, field |
+| Location / remote | Where the role is based |
+| Key responsibilities | Top 5-7 responsibilities |
+| Team / department | Context about the team |
+
+## Step 3: Parse the CV
+
+Extract and structure:
+
+| Element | What to extract |
+|---|---|
+| Name | Candidate's full name |
+| Current role + company | Most recent position |
+| Total experience | Calculated from work history |
+| Skills | Technical and soft skills (stated + inferred from experience) |
+| Education | Degrees, institutions, graduation years |
+| Certifications | Professional certifications |
+| Work history | Companies, roles, durations, key achievements |
+| Career trajectory | Progression pattern (upward, lateral, mixed) |
+| Location | Current location |
+
+## Step 4: Analyze fit
+
+Score each category and provide evidence:
+
+### Scoring rubric
+
+| Category | Weight (default) | Scoring criteria |
+|---|---|---|
+| **Skills match** | 30% | % of must-have skills present; bonus for nice-to-haves |
+| **Experience** | 25% | Years + relevance of experience to role responsibilities |
+| **Education** | 15% | Meets requirements; relevant field bonus |
+| **Career trajectory** | 15% | Growth pattern, role progression, company caliber |
+| **Role alignment** | 15% | How well current/recent work maps to this role's responsibilities |
+
+### Scoring scale
+
+| Score | Label | Meaning |
+|---|---|---|
+| 90-100 | Excellent | Exceeds requirements — strong hire signal |
+| 75-89 | Good | Meets most requirements — worth advancing |
+| 60-74 | Moderate | Meets some requirements — gaps need discussion |
+| 40-59 | Weak | Significant gaps — consider only if pipeline is thin |
+| 0-39 | Poor | Does not meet key requirements |
+
+### Red flag detection
+
+Flag these if found:
+
+| Red flag | What to look for |
+|---|---|
+| Job hopping | 3+ roles < 1 year without clear reason (contracting, startups) |
+| Gaps | Unexplained gaps > 6 months |
+| Overqualification | Significantly above the role level — flight risk |
+| Title inflation | Senior/Lead titles at very small or unknown companies with scope that doesn't match |
+| Skills mismatch | Core tech stack is completely different |
+| Stale skills | Required skills only used 5+ years ago |
+
+Important: Flag these neutrally as "items to explore in interview" — not as disqualifiers. Many have innocent explanations.
+
+## Step 5: Generate output
+
+Produce this structured report:
+
+```
+📋 CV Screening Report
+━━━━━━━━━━━━━━━━━━━━━━
+
+Candidate:    [Name]
+Role:         [JD title]
+Date:         [Today]
+
+── Overall Score: [XX]/100 — [Label] ──
+
+Recommendation: ✅ Advance / ⚠️ Hold / ❌ Pass
+
+── Category Breakdown ──
+
+| Category           | Score | Notes                           |
+|--------------------|-------|---------------------------------|
+| Skills match       | XX/30 | [brief evidence]                |
+| Experience         | XX/25 | [brief evidence]                |
+| Education          | XX/15 | [brief evidence]                |
+| Career trajectory  | XX/15 | [brief evidence]                |
+| Role alignment     | XX/15 | [brief evidence]                |
+
+── Skills Analysis ──
+
+Must-have skills:
+  ✅ Python — 5 years, used at [Company]
+  ✅ AWS — certified, 3 years
+  ❌ Kubernetes — not mentioned
+  ⚠️ SQL — mentioned but limited evidence
+
+Nice-to-have skills:
+  ✅ Terraform — 2 years
+  ❌ Go — not found
+
+── Strengths ──
+• [Strength 1 with evidence]
+• [Strength 2 with evidence]
+• [Strength 3 with evidence]
+
+── Gaps & Risks ──
+• [Gap 1 — impact assessment]
+• [Gap 2 — impact assessment]
+
+── Interview Focus Areas ──
+1. [Question topic] — to probe [specific gap or claim]
+2. [Question topic] — to verify [specific experience]
+3. [Question topic] — to assess [specific skill depth]
+
+── Red Flags ──
+• [Flag] — [neutral explanation of what to explore]
+(or "None identified")
+```
+
+### Save report
+
+Save to: `./cv-screen-<candidate-last-name>-<date>.md`
+
+If the user doesn't specify an output format, default to Markdown. Also support `--format pdf` (generate via HTML→PDF like the offer-letter skill) or `--format json` for programmatic use.
+
+## Error handling
+
+| Issue | What to do |
+|---|---|
+| CV file unreadable | Try alternative extraction methods, then ask user to provide text |
+| JD is vague / missing details | Note which categories couldn't be scored, score what's available |
+| CV has no dates | Flag as a red flag, estimate from context |
+| Multiple roles in JD | Ask user which role to screen against |