IssuePilot: Local AI-Powered Bug Triage | Hacktoberfest Week 2

A developer built IssuePilot, an open-source bug-triage tool that runs open-weight AI models locally through Ollama to convert bug descriptions, screenshots, and logs into structured reports with severity levels, reproduction steps, and recommendations. The application pairs a Spring Boot backend with a React frontend and can generate GitHub issues from the resulting reports, targeting individual developers and small teams who want to avoid paid hosted inference APIs. The developer notes that generated output should be reviewed before use and that GitHub issue creation still requires network access.

This is a submission for the Hacktoberfest Open-Source AI Challenge Week 2: Side Hustle https://dev.to/challenges/hacktoberfest-week2-2026-10-12 Bug reports are often incomplete, unclear, or missing the steps needed to reproduce a problem. Developers spend valuable time turning these reports into something actionable before they can even begin debugging. IssuePilot is a bug-triage tool that uses locally running AI models to turn bug descriptions, screenshots, and logs into structured reports. It generates: Developers can also use IssuePilot to create GitHub issues from the generated reports. I built it for individual developers and small development teams who want to reduce repetitive bug-triage work without depending on a paid, hosted AI inference API. The long-term idea is simple: spend less time formatting bug reports and more time fixing bugs. Video demo: Watch IssuePilot in action https://youtu.be/0NbLvJI QOo The demo walks through the application's bug-analysis workflow and shows how the generated report can be used. GitHub repository: See IssuePilot Source Code https://github.com/chaitanya2850/issuepilot The repository contains the application source code and setup instructions. IssuePilot combines a Java backend, a React frontend, and locally running open-weight AI models. Technology stack: The frontend sends bug descriptions or supporting evidence to the Spring Boot backend. The backend uses the appropriate local model through Ollama and returns a structured triage report. Users can review that report and create a GitHub issue when they're ready. The AI output is intended to assist developers, not replace their judgment. Generated severity levels, reproduction steps, and recommendations should be reviewed before being used. For a tool that analyzes potentially sensitive bug reports, I wanted local AI inference to be a real option rather than an afterthought. Open-weight models made it possible to experiment with text and vision capabilities without making a paid, hosted inference API a requirement. With the models downloaded, IssuePilot can perform its AI analysis locally, without sending the analysis request to an external AI inference provider. Open innovation also gives developers more control over which models they run, how they configure the application, and how they adapt it to their workflows. This matters especially for smaller teams and independent developers who may not have the budget for another recurring AI service. Local inference does not automatically make an entire application offline: GitHub issue creation still requires network access, and users should be careful about the evidence they submit. But having the AI analysis run locally is a meaningful step toward giving developers more control over their tools and data. IssuePilot is my attempt to explore how open AI models can solve a practical software-development problem, rather than being used just for a demonstration. I haven't included a DevRelay agent-session link in this submission.