Ask which crash issues affected a platform and whether they have been fixed.
Check if a bug regression appeared in a previous release before filing a new ticket.
Get a summarized, severity-ranked view of the most critical open issues.
Connect your own Jira, Firebase Crashlytics, or Splunk data to search across your real product issues.
| jsingh6/releaseradar | 0xhassaan/nn-from-scratch | 3ks/embedoc | |
|---|---|---|---|
| Stars | 0 | 0 | — |
| Language | Python | Python | Python |
| Last pushed | — | — | 2023-06-08 |
| Maintenance | — | — | Dormant |
| Setup difficulty | moderate | moderate | hard |
| Complexity | 4/5 | 4/5 | 1/5 |
| Audience | developer | developer | developer |
Figures from each repo's GitHub metadata at analysis time.
Requires your own Anthropic and GitHub API keys plus separate backend and frontend installs.
ReleaseRadar is a tool built for mobile engineering teams that lets you ask plain English questions about your software's health and get precise, cited answers back. Instead of digging through crash logs in one system, bug tickets in another, and old release notes somewhere else, you type a question and the tool pulls the relevant information together for you. The README describes the core problem this solves: when something breaks in a product used by many people, the information needed to understand what happened is usually scattered across tools that do not talk to each other. Teams end up starting from scratch every time a similar issue comes up. ReleaseRadar aims to fix that by centralizing GitHub Issues, crash events, and release notes, then letting you query them in natural language. Under the hood, it works by taking your GitHub Issues and release notes and turning them into numerical representations called embeddings, using a small open source model. When you ask a question, it combines two kinds of search: one that matches on meaning, and one that matches on exact keywords like version numbers or issue IDs, then merges the results. That combined result is handed to Claude, an AI model from Anthropic, which writes a grounded answer along with a structured summary showing severity, affected versions and platforms, and a recommendation. The frontend, built with React, displays this as a color coded card so a team can scan it quickly. The backend is written in Python with FastAPI, and data currently comes from public GitHub repositories such as Flutter and React Native, used as example datasets. The project is designed to be extended with your own data sources, including Jira, Firebase Crashlytics, and Splunk, using pluggable data connectors. Setting it up requires cloning the repository, installing Python and Node dependencies, and providing your own Anthropic and GitHub API keys. There is a live demo linked in the README. The project is released under the MIT license, meaning it can be freely used, including commercially, as long as the copyright notice is kept.
An AI tool that lets mobile engineering teams ask plain English questions across GitHub issues, crash reports, and release notes and get cited answers.
Mainly Python. The stack also includes Python, FastAPI, ChromaDB.
Use freely for any purpose, including commercial use, as long as you keep the copyright notice.
Setup difficulty is rated moderate, with roughly 30min to a first successful run.
Mainly developer.
This repo across BitVibe Labs
Verify against the repo before relying on details.