Skip to content

Latest commit

 

History

History
149 lines (112 loc) · 4.88 KB

File metadata and controls

149 lines (112 loc) · 4.88 KB

Claude Code Automation Scripts

This directory contains the automation scripts for Claude Code users who want to go beyond prompts to semi-automated job application generation.

What's Here

  • setup_career_os.py - Interactive setup wizard to build your narrative
  • apply_to_job.py - End-to-end application generation
  • validate_output.py - Quality review and validation

Requirements

  • Claude Code or Python 3.8+
  • Optional: Basic API keys for job scraping
  • 30 minutes for setup, 5 minutes per application

Quick Start

1. Initial Setup

python setup_career_os.py

This will:

  • Guide you through building your professional narrative
  • Create templates for future applications
  • Test the system with sample data

2. Apply to Jobs

python apply_to_job.py --url "https://job-posting-url.com"

Or with a saved job description:

python apply_to_job.py --file job_description.txt

3. Review Quality

python validate_output.py --application user_data/applications/latest

How It Works

Setup Process

  1. Personal Information: Basic contact details
  2. Professional Identity: How you position yourself
  3. Key Achievements: 3-5 major wins with metrics
  4. Technical Skills: Organized by category
  5. Value Proposition: What makes you unique

Application Generation

  1. Job Analysis: Strategic breakdown of what they want
  2. Resume Generation: Tailored to specific role requirements
  3. Cover Letter: Personalized narrative connecting your experience
  4. Quality Review: Structured feedback and improvement suggestions

Output Structure

Each application creates a directory with:

user_data/applications/[company]_[role]_[timestamp]/
├── job_analysis.md      # Strategic insights about the role
├── resume.md           # Tailored resume
├── cover_letter.md     # Personalized cover letter
├── quality_review.md   # Feedback and suggestions
└── metadata.json       # Application tracking data

Configuration

User Narrative

Your professional narrative is stored in:

  • user_data/narratives/main_narrative.json

This can be edited directly or regenerated with the setup script.

Application Templates

Templates are stored in:

  • templates/narrative_template.json
  • templates/resume_template.md
  • templates/cover_letter_template.md

Features

Feature Status Description
Interactive Setup ✅ Stable Guided narrative building
Job URL Scraping ⚠️ Beta Basic scraping for most sites
Resume Generation ✅ Stable Tailored formatting and content
Cover Letters ✅ Stable Personalized narrative approach
Quality Review ✅ Stable Structured feedback system
Application Tracking 🧪 Experimental Basic metadata storage

Troubleshooting

Setup Issues

  • "No narrative found": Run setup_career_os.py first
  • Permission errors: Check file permissions in user_data/
  • Python errors: Ensure Python 3.8+ is installed

Scraping Issues

  • "Error scraping job posting": Save job description to a file and use --file flag
  • Incomplete content: Some sites block automated scraping
  • Rate limiting: Wait a few minutes between attempts

Quality Issues

  • Generic output: Review and update your narrative with more specific achievements
  • Poor fit assessment: The system is designed to be honest about mismatches
  • Formatting problems: Check that templates haven't been corrupted

Tips for Best Results

Building Your Narrative

  • Be specific with metrics: "40% improvement" not "significant improvement"
  • Include context: What the achievement means for the business
  • Show progression: Demonstrate growth over time
  • Stay honest: Only include verifiable claims

Using the System

  • Review everything: Never submit without human review
  • Customize outputs: Use generated content as a starting point
  • Iterate: Update your narrative based on what works
  • Track results: Note which approaches get responses

Integration with Advanced System

This automation layer sits between the simple prompts and the full orchestrator:

  • Simple: Copy prompts to Claude/ChatGPT manually
  • Claude Code: Semi-automated with these scripts
  • Advanced: Full 6-agent orchestrator system

You can upgrade to the advanced system anytime by:

  1. Exporting your narrative from this system
  2. Setting up the advanced orchestrator
  3. Importing your data with the migration tools

Support

  • Documentation: See individual script docstrings
  • Examples: Check examples/ directory for sample outputs
  • Issues: Review error messages and logs
  • Advanced features: See advanced/ directory for full orchestrator

These scripts provide the sweet spot between manual prompts and full automation - enough automation to save time while maintaining control over quality.