This repository contains a baseline backend implementation for an automated documentation generation system. The project explores how documentation can be derived programmatically from a public GitHub repository using static analysis and repository metadata.
This is not a production-ready tool, not a full documentation generator, and not an API-powered AI system. It is a foundational backend codebase intended to demonstrate structure, feasibility, and design direction.
At its current stage, this project:
- Is backend-only
- Runs exclusively on FastAPI
- Uses no external APIs, including no LLM or AI service APIs
- Operates on public GitHub repositories only
- Produces default documentation outputs based on statically retrievable data
The system does not attempt to infer developer intent, architectural decisions, or domain-specific logic beyond what can be extracted directly from the repository.
- Clones a public GitHub repository
- Analyzes repository structure (files and folders)
- Extracts basic information from source files
- Reads Git commit history
- Generates partial documentation content
- Exports documentation in PDF and DOCX formats
The generated documentation reflects only what is directly available from the repository and version history.
It is important to be explicit about current limitations:
- Does not generate complete or polished technical documentation
- Does not use AI or LLM APIs for semantic understanding
- Does not generate architecture diagrams or design explanations
- Does not support frontend integration
- Does not run outside a FastAPI environment
- Does not produce fully structured or professional report layouts
The PDF and DOCX outputs are raw exports of extracted information, not finalized documentation reports.
Documentation automation is often discussed at a high level, but rarely shown from the ground up. This project exists to:
- Explore how documentation pipelines can be built at the backend level
- Demonstrate repository analysis using Python tooling
- Serve as a foundation for future extensions such as:
- LLM integration
- Improved document composition
- Frontend interfaces
- GitHub Actions or CI integration
This repository represents infrastructure, not a finished solution.
- Python
- FastAPI
- GitPython
- python-docx
- ReportLab
- Uvicorn
A dedicated .txt file has all the base line commands in it to run this on your local host. Follow those commands to get your output. In the terminal once the backend starts running get the local host https address and search in your browser to run via fast API.
The file will be saved to the "generated_docs" folder under "backed" folder.
This project is not in an early conceptual phase, but it is also not complete. It represents a working baseline that intentionally stops short of advanced documentation synthesis. Any claims beyond that would be inaccurate.