bug-triage

SyzBridge is a research project that adapts Linux upstream PoCs to downstream distributions. It provides rich interfaces that allow you to do a lot of cool things with Syzbot bugs

Hybrid QA bug-triage: deterministic NLP (PT) + Gemini LLM with auto-fallback | Gherkin reports for Jira/GitHub | live demo

Human-supervised AI bug triage lab with a decision ledger, duplicate radar and security-first defaults.

A spec-kit demo showing how to triage a real bug from report to merged fix using AI-assisted assess, fix, and verify agents in GitHub Copilot Chat.

Production-style AI bug triage workflow with structured classification, deterministic routing, human approval, secure diagnostics, Docker support, and CI.

AI-powered QA Intelligence Suite using CrewAI, Jira Integration, Defect Intelligence, Root Cause Analysis, Production Risk Prediction, and Regression Recommendation.

Scan Sentry hourly across a fleet, file the real bugs as GitHub issues, and hand each to Copilot for a draft PR. A human reviews every one.

A small CLI that ranks bug reports and finds possible duplicates.

RAG-based codebase intelligence system with FastAPI, ChromaDB, SentenceTransformers, Groq, background jobs, SQLite job storage, architecture explanation, bug triage, Streamlit dashboard, and tests.

MACCP: Agreement-Conditioned Conformal Prediction for Bug Triage (QRS 2026)

AI-powered bug triage system using BERT embeddings and multi-label classification

Autonomous L1 support agent built with Claude Code, Google Gemini, and Jira to triage tickets, analyze logs, and generate RCAs with code patches.

AI agent that turns messy raw bug reports into clean, structured tickets — classifies severity and priority, detects duplicates, and assigns a component. Runs offline, no API key needed.

RAG-powered multi-agent AI system for automated software defect triage, duplicate detection, log analysis, root cause analysis, and remediation using historical Apache, Eclipse, and Mozilla defect data.

AI-native product quality framework - automated VoC pipeline, LLM-as-judge bug classification, and P0 Slack alerting for developer tools

GraphCodeBERT-based embedding microservice for ELARA — converts source code into 768-dimensional semantic vectors via a FastAPI REST API, used for repository indexing and developer-code matching in ELARA's intelligent bug-assignment pipeline.

An intelligent, generic Anomaly Thought Partner skill for conversational AI agents. Guides engineers through bug reproduction, pre/post environmental state capture, quantitative risk calculation, and root-cause categorization.

Enrichissement de bug reports assisté par IA — l'IA propose, le QA arbitre. Sévérité CVSS-lite, classes IEC 62304, détection de doublons. 144 tests.

AI Bug Triage & Release Operator built on the Lemma SDK — turns messy multi-source feedback into classified, prioritized, deduplicated, reproduction-stepped issues with a human approval gate and auto-generated release notes.