GitCV

A full-stack web application that dynamically bridges GitHub's GraphQL API and Overleaf's LaTeX compilation engine. GitCV automates the generation of a professional, typeset curriculum vitae by extracting a user's repository data, processing it through an LLM for professional phrasing, and injecting it into a parameterized TeX template.

System Architecture & Data Flow

1. Authentication: OAuth 2.0 flow via GitHub to provision a short-lived user access token.
2. Data Extraction: Queries the GitHub GraphQL API to batch-fetch user bios, pinned repositories, and language statistics in a single network request.
3. LLM Transformation (BYOK): Intercepts raw repository descriptions and routes them to an LLM (OpenAI/Gemini) using user-provided API keys. The LLM parses technical git-commits into impact-driven resume bullet points.
4. LaTeX Sanitization: Passes the LLM output through a strict sanitization layer to safely escape TeX special characters (e.g., &, %, _) to prevent compiler failures.
5. Template Injection: Binds the sanitized JSON payload to a base .tex.jinja template using the Jinja2 engine.
6. Packaging & Routing: Bundles the generated .tex file into an in-memory .zip archive, hosts it temporarily, and redirects the client to Overleaf via the snip_uri protocol for instant remote compilation.

Tech Stack

• Backend: Python 3.10+, FastAPI, Uvicorn
• Templating: Jinja2
• Integrations: GitHub GraphQL API, OpenAI/Gemini REST APIs, Overleaf URL Import API

Local Development Setup

Prerequisites

• Python 3.10+
• A registered GitHub OAuth Application (Requires CLIENT_ID and CLIENT_SECRET)
• An OpenAI or Gemini API key (for local LLM integration testing)

Installation & Execution

1. Clone the repository:

   git clone [https://github.com/YOUR_USERNAME/gitcv.git](https://github.com/YOUR_USERNAME/gitcv.git)
   cd gitcv/backend

2. Configure environment variables:

   cp .env.example .env

   Open .env and populate it with your GitHub OAuth credentials and local testing keys. Ensure .env is listed in your .gitignore.

3. Initialize the virtual environment:

   python -m venv venv
   source venv/bin/activate  # On Windows: venv\Scripts\activate

4. Install dependencies:

   pip install -r requirements.txt

5. Start the ASGI development server:

   uvicorn app.main:app --reload

   The API will be available at http://127.0.0.1:8000. Access http://127.0.0.1:8000/docs for the interactive Swagger UI.

Development Roadmap

Status: MVP in progress. Core API scaffolding complete.

• Phase 0: Project Setup & Repo Scaffolding
• Phase 1: GitHub OAuth flow & token exchange
• Phase 2: GitHub GraphQL data fetching & edge-case testing
• Phase 3: LLM Integration & prompt engineering for resume generation
• Phase 4: LaTeX Sanitization layer & unit testing
• Phase 5: Jinja2 LaTeX templating & local TeX compilation checks
• Phase 6: In-memory Zip creation & Overleaf snip_uri routing
• Phase 7: End-to-End Integration & Frontend UI build
• Phase 8: Security hardening, rate limiting, and public release

License

Distributed under the Apache 2.0 License.