Skip to content

TAILOR_FROM_URL_INTEGRATION.md

Latest commit

 

History

History
339 lines (260 loc) · 8.02 KB

TAILOR_FROM_URL_INTEGRATION.md

File metadata and controls

339 lines (260 loc) · 8.02 KB

Tailor Resume from URL - Integration Guide

Overview

The tailor_from_url.py script provides a simple, unified workflow for:

  1. Fetching job listings from URLs
  2. Tailoring resumes to job descriptions
  3. Generating HTML/DOCX output

This is the simplest way to go from job URL → tailored resume.

Quick Start

Basic Usage

# Tailor Master Resume to job URL
python src/tailor_from_url.py \
  --url "https://example.com/job-posting" \
  --resume "Master Resume" \
  --out out/tailored.html

# With DOCX export
python src/tailor_from_url.py \
  --url "https://example.com/job-posting" \
  --resume "Master Resume" \
  --out out/tailored.html \
  --docx

# With RAG enhancement
python src/tailor_from_url.py \
  --url "https://example.com/job-posting" \
  --resume "Master Resume" \
  --out out/tailored.html \
  --use-rag \
  --theme modern

Features

✅ What It Does

  1. Fetches Job Listing

    • Downloads HTML from URL
    • Extracts job title, company, location, description
    • Saves as markdown to data/job_listings/
    • Auto-updates data/job_listings/index.json

  2. Tailors Resume

    • Loads your resume (by name or file path)
    • Extracts keywords from job description
    • Scores and selects relevant experience bullets
    • Optionally uses RAG for context-aware selection

  3. Generates Output

    • Creates professional HTML resume
    • Optionally exports to DOCX
    • Supports multiple themes

🎯 Supported Themes

  • professional (default) - Clean, corporate look
  • modern - Contemporary design
  • executive - Premium appearance
  • creative - Unique, artistic style

Command Reference

Required Arguments

Argument Description Example
--url Job listing URL https://example.com/job
--resume Resume name or file path "Master Resume" or data/master_resume.json
--out Output HTML file path out/tailored.html

Optional Arguments

Argument Description Default
--theme HTML theme professional
--docx Also export to DOCX False
--use-rag Use RAG for tailoring False
--use-llm-rewriting Use LLM for rewriting False
--vector-store RAG vector store path data/rag/vector_store.json

Examples

Example 1: Simple Tailoring

python src/tailor_from_url.py \
  --url "https://www.linkedin.com/jobs/view/123456789" \
  --resume "Master Resume" \
  --out out/linkedin_job.html

Output:

  • out/linkedin_job.html - Tailored resume
  • data/job_listings/LinkedIn_Job_Title.md - Saved job listing
  • data/job_listings/index.json - Updated index

Example 2: With DOCX Export

python src/tailor_from_url.py \
  --url "https://example.com/careers/senior-engineer" \
  --resume "Ford" \
  --out out/ford_senior_engineer.html \
  --docx \
  --theme modern

Output:

  • out/ford_senior_engineer.html - HTML resume
  • out/ford_senior_engineer.docx - DOCX resume
  • data/job_listings/Senior_Engineer.md - Saved job listing

Example 3: With RAG Enhancement

python src/tailor_from_url.py \
  --url "https://example.com/devops-role" \
  --resume "Master Resume" \
  --out out/devops_tailored.html \
  --use-rag \
  --theme executive

Features:

  • Retrieves relevant experiences from RAG vector store
  • Scores bullets based on job requirements
  • Selects most relevant experiences
  • Generates tailored resume

Workflow Diagram

User provides URL
        ↓
Fetch job listing from URL
        ↓
Extract job title, company, location, description
        ↓
Save to data/job_listings/
        ↓
Update data/job_listings/index.json
        ↓
Load resume (by name or path)
        ↓
Extract keywords from job description
        ↓
[Optional] Retrieve RAG context
        ↓
Score and select relevant bullets
        ↓
Generate HTML resume
        ↓
[Optional] Export to DOCX
        ↓
Output files ready

Resume Lookup

The script supports multiple ways to specify your resume:

By File Path

--resume "data/master_resume.json"
--resume "data/resumes/abc123.json"

By Resume Name

--resume "Master Resume"
--resume "Ford"
--resume "Sidney_Jones_Engineering_Manager"

The system automatically looks up the resume by name in data/resumes/index.json.

Output Files

Generated Files

  1. HTML Resume

    • Location: out/tailored.html (or specified path)
    • Format: Professional HTML with embedded CSS
    • Themes: Professional, Modern, Executive, Creative

  2. DOCX Resume (if --docx flag used)

    • Location: out/tailored.docx
    • Format: Microsoft Word document
    • Fully editable

  3. Job Listing (auto-saved)

    • Location: data/job_listings/Job_Title.md
    • Format: Markdown
    • Contains: Title, company, location, description

  4. Index Update (auto-updated)

    • Location: data/job_listings/index.json
    • Contains: Metadata for all fetched job listings
    • Fields: ID, title, company, location, file, created_at

Integration with Agent

The agent can use this script to tailor resumes:

User: "Tailor my resume for this job: https://example.com/job"

Agent: "I'll tailor your resume to that job posting.

run: python src/tailor_from_url.py --url "https://example.com/job" --resume "Master Resume" --out out/tailored.html --docx

[After execution]
✅ Resume tailored successfully!
   - HTML: out/tailored.html
   - DOCX: out/tailored.docx
   - Job listing saved to: data/job_listings/Job_Title.md

Troubleshooting

Issue: "ModuleNotFoundError: No module named 'fetch_job_listing'"

Solution: Make sure you're running from the repository root:

cd c:\source\repos\bpm\internal\agentic-resume-tailor
python src/tailor_from_url.py --url "..." --resume "..." --out "..."

Issue: "Resume not found"

Solution: Check that the resume name exists:

# List available resumes
python -c "import json; print(json.load(open('data/resumes/index.json')))"

Issue: "Job listing fetch failed"

Solution: Some sites have anti-bot protection. Try:

  1. Using a different URL
  2. Checking if the URL is publicly accessible
  3. Using a local HTML file for testing

Issue: "RAG context not found"

Solution: Make sure the vector store exists:

ls -la data/rag/vector_store.json

If missing, build it first:

python src/rag/rag_indexer.py

Testing

Run the integration test:

python scripts/test_tailor_from_url.py

Expected output:

✅ ALL TESTS PASSED

Performance

  • Fetch job listing: ~2-5 seconds
  • Extract keywords: ~0.5 seconds
  • Tailor resume: ~1-2 seconds
  • Generate HTML: ~1-2 seconds
  • Export DOCX: ~2-3 seconds

Total time: ~7-15 seconds (depending on options)

Advanced Usage

Custom Vector Store

python src/tailor_from_url.py \
  --url "https://example.com/job" \
  --resume "Master Resume" \
  --out out/tailored.html \
  --use-rag \
  --vector-store "data/rag/custom_store.json"

LLM-Enhanced Rewriting

python src/tailor_from_url.py \
  --url "https://example.com/job" \
  --resume "Master Resume" \
  --out out/tailored.html \
  --use-rag \
  --use-llm-rewriting

Requires:

  • RAG enabled (--use-rag)
  • LLM API key configured
  • Vector store populated

Architecture

tailor_from_url.py
├── fetch_job_listing()      # Fetch from URL
├── ingest_jd()              # Parse job description
├── extract_keywords()       # Extract keywords
├── retrieve_rag_context()   # [Optional] RAG retrieval
├── load_resume()            # Load resume
├── select_and_rewrite()     # Tailor bullets
├── generate_html_resume()   # Generate HTML
└── generate_docx_from_html() # [Optional] Export DOCX

See Also

Status: ✅ Production Ready
Last Updated: 2025-10-26