/screen-resume Command

AI-powered resume screening tool for quickly evaluating candidate fit before scheduling interviews.

Usage

/screen-resume [/path/to/resume or /path/to/directory]

[role]
[seniority level]
[must-have skills/requirements]

[optional: disqualifiers or red flags]

Required Parameters

• Resume path: Path to candidate's PDF resume OR directory containing multiple resumes
• Role: Position being hired for (e.g., Mobile Developer, Backend Engineer, Full Stack Developer)
• Seniority level: Junior, Mid-level (Pleno), Senior, Lead, Principal
• Must-have requirements: Core skills/experience that are non-negotiable

Optional Parameters

• Disqualifiers: Specific red flags or dealbreakers to watch for
• Additional context: Team needs, project specifics, culture fit considerations

Command Pipeline

Step 1: Input Validation

• Validate all required parameters are provided
• Check if path is file or directory
• If directory, collect all PDF files
• Prompt for missing required fields

Step 2: File Processing

• For single file: Process immediately
• For directory: Batch process all PDFs
• Convert each PDF to markdown for analysis
• Create temporary workspace for screening

Step 3: Resume Screening Analysis

Perform rapid assessment focusing on:

• Role Match Score (0-100%): How well experience matches the role
• Seniority Alignment: Does experience match the expected level?
• Must-Have Skills Coverage: Checklist of required skills
• Red Flags: Any disqualifiers or concerns
• Green Flags: Exceptional qualifications or experiences

Step 4: Scoring & Decision

Generate screening verdict:

• STRONG MATCH (80-100%): Definitely interview
• GOOD MATCH (60-79%): Worth interviewing
• MAYBE (40-59%): Consider if no better candidates
• WEAK MATCH (20-39%): Likely not a fit
• NO MATCH (0-19%): Does not meet requirements

Step 5: Output Generation

Create structured screening report:

• Executive summary (3-5 bullet points)
• Detailed scoring breakdown
• Specific concerns or highlights
• Recommended interview focus areas (if proceeding)
• Salary expectations analysis (if mentioned)

Step 6: Follow-up Actions

Based on screening results:

• Strong/Good Match: Prompt to run /generate-interview
• Maybe: Suggest specific areas to probe further
• Weak/No Match: Provide rejection template or feedback
• For batch processing: Generate ranked candidate list

Output Structure

Single Resume

/Users/nagawa/v2t/resume-screening/
└── [timestamp]-screening/
    ├── [candidate-name]-resume.pdf (original)
    ├── [candidate-name]-resume.md (converted)
    └── [candidate-name]-screening-report.md

Batch Processing (Directory)

/Users/nagawa/v2t/resume-screening/
└── [timestamp]-batch-screening/
    ├── batch-summary.md (ranked list of all candidates)
    ├── strong-matches/
    │   └── [candidate-name]-screening.md
    ├── good-matches/
    │   └── [candidate-name]-screening.md
    ├── maybe/
    │   └── [candidate-name]-screening.md
    └── rejected/
        └── [candidate-name]-screening.md

Screening Report Format

# Resume Screening Report - [Candidate Name]
**Date:** [Date]
**Role:** [Position] - [Seniority]
**Verdict:** [STRONG MATCH/GOOD MATCH/MAYBE/WEAK MATCH/NO MATCH]

## Executive Summary
✅ **Score: [X]%**
- [Key qualification 1]
- [Key qualification 2]
- [Main concern if any]

## Detailed Analysis

### Role Match (X/100)
**Relevant Experience:** [Years] years
- [Specific relevant role/project]
- [Another relevant experience]

### Required Skills Coverage
✅ [Skill 1] - Strong evidence
✅ [Skill 2] - Some experience
❌ [Skill 3] - Not evident
⚠️ [Skill 4] - Unclear/needs verification

### Seniority Assessment
**Expected:** [Level]
**Actual:** [Assessment]
**Justification:** [Why they match or don't match the level]

### Red Flags
🚩 [Concern 1]
🚩 [Concern 2]

### Green Flags
💚 [Exceptional qualification 1]
💚 [Exceptional qualification 2]

## Recommendation
[PROCEED TO INTERVIEW / CONSIDER WITH CAUTION / REJECT]

### If Interviewing, Focus On:
1. [Area to probe deeply]
2. [Skill to verify]
3. [Experience to explore]

### Interview Type Suggestion:
- [Recommended interview format based on profile]
- [Duration recommendation]

---
💡 **Quick Action:** Run `/generate-interview [resume-path]` to create interview questions

Batch Processing Report

# Batch Resume Screening Summary
**Date:** [Date]
**Role:** [Position] - [Seniority]
**Total Screened:** [X] candidates

## Ranking

### 🏆 Strong Matches ([X] candidates)
1. **[Name]** - [Score]% - [Key strength]
2. **[Name]** - [Score]% - [Key strength]

### ✅ Good Matches ([X] candidates)
1. **[Name]** - [Score]% - [Note]
2. **[Name]** - [Score]% - [Note]

### 🤔 Maybe ([X] candidates)
1. **[Name]** - [Score]% - [Main concern]

### ❌ Rejected ([X] candidates)
1. **[Name]** - [Score]% - [Disqualifier]

## Recommendations
1. **Immediate interviews:** [Names]
2. **Second tier:** [Names]
3. **Only if needed:** [Names]

## Next Steps
Run `/generate-interview` for top candidates:
- `/generate-interview /path/to/[candidate1].pdf`
- `/generate-interview /path/to/[candidate2].pdf`

Screening Criteria Framework

Technical Skills Assessment

• Direct match: Exact technology/framework experience
• Transferable: Similar technologies (e.g., React → React Native)
• Foundation: Underlying concepts (e.g., any frontend → specific framework)
• Missing: No evidence of skill or related experience

Experience Level Mapping

• Junior: 0-2 years, learning mindset, basic projects
• Mid/Pleno: 2-5 years, independent work, some complexity
• Senior: 5+ years, leadership, architecture, mentoring
• Lead/Principal: 8+ years, strategy, cross-team impact

Red Flag Detection

• Frequent job hopping (< 1 year stays without explanation)
• Technology mismatch (e.g., only backend for frontend role)
• Overqualification (senior applying for junior role)
• Geographic/timezone incompatibility
• Salary expectation mismatch

Green Flag Recognition

• Exact industry experience
• Open source contributions
• Relevant certifications
• Growth trajectory
• Cultural fit indicators
• Side projects showing passion

Integration with /generate-interview

After screening completes with STRONG or GOOD match:

Screening complete! Candidate is a STRONG MATCH (85%).

Would you like to generate interview questions? (y/n)
> y

Proceeding to /generate-interview...
[Automatically passes candidate info and screening insights]

/generate-interview /path/to/resume.pdf
[Pre-fills candidate name from screening]
[Pre-fills role and seniority]
[Includes screening focus areas as context]

Command Implementation

When invoked, Claude Code will:

1. Parse Input: Extract path and requirements
2. Detect Type: Single file vs directory
3. Process Files: Convert PDFs to markdown
4. Quick Analysis: Rapid pattern matching against requirements
5. Score & Rank: Generate match percentages
6. Generate Reports: Create screening summaries
7. Prompt Next: Offer to proceed with interview generation
8. Store Context: Save screening patterns for learning

Screening Speed Optimizations

• Parallel Processing: For directories, screen multiple resumes simultaneously
• Quick Scan: Focus on keywords and patterns first
• Smart Extraction: Pull only relevant sections for deep analysis
• Caching: Store converted markdown for re-screening
• Templates: Pre-built assessment criteria by role

Example Execution

Single File

User: /screen-resume /Downloads/john_doe_resume.pdf

Backend Engineer
Senior
Python, Django, PostgreSQL, Docker, AWS

No consultancy experience, must have product company background

Claude:
🔍 Screening resume for Backend Engineer (Senior)...
✅ Resume converted and analyzed

## John Doe - STRONG MATCH (82%)

### Quick Summary:
✅ 7 years backend experience
✅ Python/Django expert (5 years)
✅ AWS certified, Docker proficient
✅ Product company background (Spotify, Uber)
⚠️ PostgreSQL experience limited (mostly MySQL)

### Recommendation: PROCEED TO INTERVIEW
Focus areas: PostgreSQL specifics, system design, scaling challenges

Generate interview questions now? (y/n)

Directory Batch

User: /screen-resume /Downloads/backend_candidates/

Backend Engineer
Senior
Python, Django, PostgreSQL, Docker, AWS

Claude:
📁 Found 12 resumes to screen...
⏳ Processing batch screening...

## Screening Complete!

### Results Summary:
🏆 **Strong Matches (3)**
1. John Doe - 82% - Spotify/Uber experience
2. Maria Silva - 79% - Django expert, AWS certified
3. Carlos Santos - 77% - Strong PostgreSQL, Docker

✅ **Good Matches (2)**
4. Ana Costa - 65% - Good skills, less seniority
5. Pedro Lima - 61% - Solid backend, learning Python

🤔 **Maybe (2)**
❌ **Rejected (5)** - Wrong tech stack or junior level

View detailed reports in: /Users/nagawa/v2t/resume-screening/[timestamp]-batch-screening/

Generate interviews for top 3? (y/n)

Error Handling

• Missing file: Prompt for valid path
• Invalid PDF: Try alternative conversion methods
• Corrupted file: Skip and note in batch report
• Missing requirements: Interactive prompt for criteria
• Large batches: Process in chunks with progress updates

Performance Targets

• Single resume: < 10 seconds
• Batch of 10: < 60 seconds
• Accuracy goal: 90% agreement with human screening
• False negative rate: < 5% (avoid rejecting good candidates)

Context Learning

Store screening patterns in ~/.claude/screening-patterns/:

• Successful hire patterns by role
• Common red flags that proved accurate
• Green flags that correlate with performance
• Skill transferability mappings

Future Enhancements

• Auto-detect role from job description
• LinkedIn profile integration
• GitHub profile analysis for developers
• Portfolio review for designers
• Salary benchmark comparison
• Diversity and inclusion metrics
• ATS integration for bulk processing