GitHired — GitHub-based AI CV & Cover Letter Generator

The following is create by https://github.com/Herc-Ch/readme-generator

Generate polished, LaTeX-ready CVs and cover letters using your existing CV, your GitHub repositories (top-level READMEs), and live job postings. Powered by LangChain, OpenAI, PyGithub, Tavily, and optional local PDF compilation.

✨ Ideal for AI Engineer / RAG Engineer / Prompt Engineer profiles (configurable).

---

Features

• Pulls top-level README.md from all your GitHub repositories (private or public with token).
• Rewrites your existing LaTeX CV or designs a new modern template.
• Generates tailored cover letters from a job URL by crawling the posting.
• Produces clean, compilable LaTeX (.tex) output; optional local PDF via xelatex.
• Uses LangChain prompt templates for reproducibility and clarity.
• Tracing support via LangSmith (optional).

---

Installation

Prerequisites:

• Python 3.10+
• Git (for cloning repos)
• LaTeX distribution (for optional PDF): TeX Live / MiKTeX / MacTeX

Create and activate a virtual environment, then install dependencies:

# Clone your repo (or download the source)
git clone <your-repo-url>
cd <your-repo-folder>

# Create and activate a virtual environment
python -m venv .venv
# Windows
.venv\Scripts\activate
# macOS/Linux
source .venv/bin/activate

# Install Python deps
pip install -r requirements.txt

Set up your environment variables in a .env file at the project root:

# .env (keys only; provide your own values)
GITHUB_TOKEN=
OPENAI_API_KEY=
LANGSMITH_API_KEY=
LANGSMITH_TRACING=
TAVILY_API_KEY=

Notes:

• GITHUB_TOKEN must have permission to read your repositories (e.g., repo scope for private repos).
• TAVILY_API_KEY is required for job page crawling.
• LANGSMITH_* are optional for tracing.

---

Usage

The default entry point is src/main.py, which:

• Generates/rewrites your CV.
• Creates a cover letter tailored to the JOB_URL constant inside src/main.py.
• Writes outputs into the outputs/ directory.

Quick start:

# Run the pipeline (edit JOB_URL in src/main.py first)
python src/main.py

Customize behavior:

• To regenerate the CV with GitHub context and updated content, create_cv(force_update=True) is used by default in main.py.
• To reuse your base CV without regeneration, set force_update=False in code.
• To target a different role/focus, update the fields arguments in:
  • src/cv_templates.py (TEMPLATE_REWRITE / TEMPLATE_NEW)
  • src/cover_letter/cover_letter_templates.py
  • or pass different values via summarize_documents when integrating.

Programmatic usage:

python - << 'PY'
from src.main import create_cv, create_cover_letter
from pathlib import Path

# 1) Generate an updated CV (or set force_update=False to reuse base CV in data/)
cv_path = create_cv(force_update=True)

# 2) Generate a cover letter for a specific job posting
job_url = "https://example.com/job-posting"
create_cover_letter(job_url, generate=True, cv_path=cv_path)
PY

Outputs:

• outputs/cv_updated.tex
• outputs/cover_letter_updated.tex

Optional: compile to PDF locally (requires LaTeX installed):

python - << 'PY'
from src.utils.pdf_compiler import generate_pdf_local

# Compile CV
generate_pdf_local("outputs/cv_updated.tex")

# Compile Cover Letter
generate_pdf_local("outputs/cover_letter_updated.tex")
PY

---

Environment Variables

Provide the following keys in .env:

• GITHUB_TOKEN
• OPENAI_API_KEY
• LANGSMITH_API_KEY
• LANGSMITH_TRACING
• TAVILY_API_KEY

No secret values are included in this README. Please supply your own.

---

Architecture

flowchart LR
  A["User LaTeX CV (.tex)"]
  B["GitHub API (READMEs)"]
  C["Tavily Crawl (Job Page)"]
  D["LangChain Orchestrator"]
  E["OpenAI Chat (LLM)"]
  F["Templates & Prompts"]
  G["Outputs (.tex)"]
  H["Local PDF Compiler (XeLaTeX)"]
  I["LangSmith (Optional Tracing)"]

  subgraph EXT["External Services"]
    B
    C
    E
    I
  end

  A -- "Context" --> D
  B -- "Top-level README.md" --> D
  C -- "Job description text" --> D
  F -- "Prompt + Styling" --> D
  D -- "LLM calls" --> E
  E -- "Responses" --> D
  D -- "Generate" --> G
  G -- "Compile (Optional)" --> H
  D -- "Telemetry" --> I

Loading

---

Project Structure

.
├─ .env                         # Your local secrets (not committed)
├─ requirements.txt             # Python dependencies
├─ src/
│  ├─ main.py                   # Entry point: CV + Cover Letter generation
│  ├─ cv_templates.py           # CV prompt templates (rewrite/new)
│  ├─ git_loader.py             # GitHub auth + README loader
│  ├─ cover_letter/
│  │  ├─ cover_letter_templates.py  # Cover letter prompt templates
│  │  └─ web_scraper.py             # Job page crawler (Tavily)
│  └─ utils/
│     ├─ file_paths.py          # Path utilities and defaults
│     ├─ loaders.py             # Document loaders (CV, cover letter, job page)
│     └─ pdf_compiler.py        # Optional LaTeX → PDF helper
├─ data/
│  ├─ cv_private.tex            # Your base CV (optional)
│  ├─ example_cv.tex            # Fallback example CV
│  ├─ cover_letter_private.tex  # Your base cover letter (optional)
│  └─ example_cover_letter.tex  # Fallback example cover letter
├─ outputs/
│  ├─ cv_updated.tex
│  └─ cover_letter_updated.tex
└─ repos_cache/                  # Cloned repos for README ingestion

---

Notes and Tips

• The system adds a subtle TikZ watermark credit line by default; ensure your LaTeX environment supports tikz and hyperref.
• Default LLM model is "gpt-5" in code; update to your available OpenAI model as needed in src/main.py (summarize_documents).
• Only top-level README.md files are ingested for each repository.
• If you hit rate limits or auth errors, verify your tokens and scopes.
• Overleaf: The generated LaTeX is standard and should compile without custom .cls/.sty files.

Enjoy your streamlined job applications 🚀