Every abandoned project has a story. Let yours speak again.
This is a submission for the Hacktoberfest Open-Source AI Challenge Week 4: Ghost in the Machine
What I Built
RE:VENANT is a local AI-powered repository archaeology and resurrection tool that gives abandoned software projects a second life.
Every developer has probably started a project that never reached completion. Over time, repositories become forgotten collections of unfinished features, TODO comments, incomplete implementations, outdated documentation, and development decisions.
But what if those repositories could tell their own stories?
That's the idea behind RE.
Users can investigate a local Git repository and interact with an AI-powered representation of its history, almost like communicating with the ghost of the project itself.
Its core capabilities include:
Repository Séance: Investigates a local Git repository, its structure, source files, and development history.
Digital Autopsy: Identifies unfinished tasks, incomplete documentation, potential technical problems, and development activity.
Talk to the Ghost: Allows developers to ask questions about the project's history and unfinished work.
Resurrection Roadmap: Generates a prioritized development plan to help bring the project back to life.
Ghost Mode: Provides an immersive, terminal-inspired interface for interacting with the repository's AI.
The central idea is simple: instead of letting abandoned software remain forgotten, use AI to understand what is already there and help developers continue where they left off.
Code
GitHub Repository:
The project is intended to be developed as a local-first, open-source developer tool.
How I Built It
I chose a local-first architecture so that the application's core intelligence does not depend on external AI APIs.
Technology Stack
Frontend: Next.js, React, TypeScript
Styling: Tailwind CSS
AI Model: Gemma 3 4B
Local Inference: Ollama
Repository Analysis: Git CLI and Node.js
Backend: Next.js API routes
Data Handling: Local filesystem and structured repository analysis
How It Works
The application follows a straightforward pipeline:
The user selects a local Git repository.
The repository investigation engine extracts relevant metadata, commit history, source files, and unfinished tasks.
The extracted information is organized into structured evidence.
Gemma 3 4B processes that evidence through Ollama running locally.
The AI generates an investigation report, responds to questions, and creates a practical resurrection roadmap.
One important design decision was to distinguish actual repository evidence from AI-generated interpretations.
For example, the system should never claim to know why a developer abandoned a project simply because its last commit was several months ago.
Instead, it identifies observable facts and presents possible explanations as hypotheses.
Why Does Open Innovation Matter?
For RE, open innovation is not just a technology choice. It is central to the problem the project attempts to solve.
Software repositories can contain years of personal effort, unfinished ideas, private source code, and intellectual work.
Sending all of that information to an external AI service should not be the only way to benefit from AI-assisted development.
By using Gemma 3 4B through Ollama, RE can perform its AI processing locally without depending on a proprietary cloud inference API.
This provides several advantages:
Privacy: Repository contents can remain on the developer's machine rather than being uploaded to a third-party AI provider.
Local inference: Once the model is installed, AI functionality can operate without an internet connection.
Model flexibility: The underlying model can be replaced with another compatible local model.
Accessibility: There are no per-request cloud AI API charges.
Control: Developers have greater control over how their repository data is processed.
The project uses an open-weight model rather than claiming that every component of the model ecosystem is fully open-source.
I believe open innovation matters because useful AI tools should be accessible, adaptable, and not unnecessarily dependent on centralized services.
For this particular project, the ability to investigate private code locally is especially valuable.
The Spirit of It
There is something interesting about abandoned software.
A repository is more than a collection of files. It contains traces of someone's curiosity, effort, experiments, decisions, and unfinished ambitions.
Sometimes a project is not abandoned because the idea was bad. Sometimes its creator simply ran out of time, motivation, or resources.
RE explores the idea that these projects deserve another look.
Rather than treating unfinished code as something useless, I wanted to build a tool that helps developers rediscover what they started and gives them a practical way to continue.
Because even forgotten code deserves a second chance.
Every abandoned project has a story. Let yours speak again.
Top comments (0)