Local Job Pipeline

Local Job Pipeline is a local-first job application pipeline for collecting public job postings, scoring them, generating local resume, cover-letter, and answer-pack drafts, and tracking manual applications.

This public repository contains code structure, tests, policies, launcher scripts, and fake examples only. Private candidate data is intentionally excluded.

Safety Boundary

This project is designed for manual, local workflows.

• No auto-apply.
• No auto-submit.
• No login automation.
• No captcha bypass.
• No external LLM or API requirement for local generation.
• Manual review is required before any application material is used.
• Sensitive fields must be answered manually.

Generated drafts are only drafts. Review every factual claim against your own local profile and the job posting before using it.

Directory Layout

• config/: public-safe configuration defaults and local workflow settings.
• resources/: tracked public examples, templates, policies, and job taxonomy files.
• local_resources/: ignored private profile files that you create locally.
• generated/: ignored generated resumes, cover letters, answer packs, reports, and workspaces.
• data/: ignored local databases, raw job pulls, processed outputs, and logs.
• job_pipeline/: Python package source.
• tests/: unit tests that run without private data.

Resource Setup

Create your local private candidate profile from the public example:

New-Item -ItemType Directory -Force local_resources\candidate
Copy-Item resources\candidate\master_profile.example.yaml local_resources\candidate\master_profile.yaml

Then edit local_resources/candidate/master_profile.yaml with your own facts. Do not edit tracked example files with real resume facts.

Create role-specific private profiles under:

local_resources/role_profiles/*.yaml

Tracked role examples live under resources/role_profiles/*.example.yaml and contain fake data only.

Generated Layout

Application workspaces are written under:

generated/YYYYMMDD/company/role__jobid/

Formal resume and cover-letter files are written at the workspace root, for example:

generated/20260630/example_company/example_role__job_123/Candidate_Name__Example_Company__Example_Role__Resume.docx
generated/20260630/example_company/example_role__job_123/Candidate_Name__Example_Company__Example_Role__Cover_Letter.pdf

Source snapshots and review files are stored inside the same workspace so each draft can be audited before use.

Private Data Warning

Never commit:

• local_resources/
• data/
• generated/
• SQLite or DB files
• logs
• real resume facts
• real candidate profile YAML
• generated application materials
• credentials, tokens, or .env files

Only fake examples should be tracked in this public repository.

Quick Start

Install dependencies in your preferred Python environment, then run tests:

python -m pip install -r requirements.txt
python -m unittest discover -s tests

Start the local dashboard:

streamlit run job_pipeline/dashboard.py

You can also use the included local launcher scripts when running on Windows:

.\start_local_service.ps1

Public Version Note

This public repo mirrors the private project architecture while excluding private candidate data, generated applications, databases, logs, backups, resume files, and private Git history.