ARIA v1.6.0 — Repository Intelligence & Reliability

ARIA v1.6.0 — Repository Intelligence & Reliability

ARIA v1.6.0 builds on the v1.5.x production-hardening line with improved
repository grounding, resilient LLM orchestration, self-hosted persistence,
and CI reliability.

Highlights

• Improved Repository Chat grounding and deterministic symbol resolution
• Multi-provider LLM failover:
  Gemini → DeepSeek → Llama → MiniMax
• Circuit-breaker and token-aware streaming failover
• Improved retrieval grounding and citation verification
• ONNX INT8 embedding pipeline with PyTorch fallback
• Fixed SQLite persistence for self-hosted Docker deployments
• Improved self-hosted Docker deployment topology
• Frontend CI Node.js compatibility fixes
• Improved Linux CI and frontend typecheck reliability
• Expanded regression, retrieval, grounding, provider, and integration tests
• Documentation synchronized with the current architecture

Validation

• Backend test suite: 0 failures
• Frontend: 393/393 tests passing
• TypeScript: clean
• Ruff: clean
• GitHub Actions: all required jobs passing
• Container build: passing
• Self-hosted SQLite persistence: verified across container restart

Deployment

ARIA is designed for self-hosted deployment using Docker Compose.

The canonical deployment configuration is docker-compose.yml.

Upgrade Notes

This release is recommended as the baseline for the latest repository
intelligence, grounding, reliability, and self-hosted persistence improvements.

What's Changed

• fix: secure Vercel API proxy and async analysis flow by @VarshithReddy2006 in #15
• fix: use TypeScript Vercel serverless entrypoint by @VarshithReddy2006 in #16

Full Changelog: v1.5.1...v1.6.0

ARIA v1.5.1 — Production Hardening

ARIA v1.5.1 — Production Hardening

ARIA v1.5.1 is a production-hardening patch release for the v1.5 architecture, focused on Azure deployment reliability and asynchronous repository analysis.

🚀 Highlights

• Hardened asynchronous repository analysis job dispatch
• Prevented false 202 Accepted responses when job dispatch fails
• Added shared persistent job state through JOB_STATE_DIR
• Fixed Azure Container Apps revision host validation
• Improved ALLOWED_HOSTS parsing and Azure hostname handling
• Added worker progress and Azure queue observability
• Added regression coverage for dispatch failures and job-state persistence
• Added automated staging end-to-end deployment verification

☁️ Production Validation

The complete production pipeline was validated:

API
 ↓
Azure Storage Queue
 ↓
ARIA Worker
 ↓
Repository Analysis
 ↓
Shared Job State
 ↓
API Polling
 ↓
Completed Analysis

## What's Changed
* feat: implement Phase 2 intelligence suite (PR Intelligence, Architecture Drift, Dead Code Analysis) by @VarshithReddy2006 in https://github.com/VarshithReddy2006/ARIA/pull/8
* feat(ui): overhaul frontend with design system, accessibility, and dashboard improvements by @VarshithReddy2006 in https://github.com/VarshithReddy2006/ARIA/pull/9
* Fix backend-lint-and-test CI failure caused by Tree-Sitter package incompatibility by @VarshithReddy2006 with @Copilot in https://github.com/VarshithReddy2006/ARIA/pull/11
* Revise README content and add live demo section by @VarshithReddy2006 in https://github.com/VarshithReddy2006/ARIA/pull/12
* fix: resolve Pydantic compatibility for MCP SDK by @VarshithReddy2006 in https://github.com/VarshithReddy2006/ARIA/pull/13
* Release/aria v1.5.0 by @VarshithReddy2006 in https://github.com/VarshithReddy2006/ARIA/pull/14

## New Contributors
* @VarshithReddy2006 with @Copilot made their first contribution in https://github.com/VarshithReddy2006/ARIA/pull/11

**Full Changelog**: https://github.com/VarshithReddy2006/ARIA/compare/v3.0.0...v1.5.1

# 🚀 Repository Intelligence Agent v1.5.0 -Production Release

We're excited to announce Repository Intelligence Agent v1.5.0, the first production-ready release of an AI-native platform for understanding, analyzing, and navigating complex software repositories.

Built around the Repository Intelligence Architecture (RIA), this release combines static analysis, semantic search, repository intelligence, and the Model Context Protocol (MCP) to help developers, AI coding assistants, and engineering teams understand unfamiliar codebases faster.

---

🌟 What's New

🤖 Native Model Context Protocol (MCP)

Repository Intelligence Agent now exposes production-ready MCP servers, enabling seamless integration with AI coding assistants such as Cursor, Claude Desktop, VS Code MCP clients, and other MCP-compatible tools.

🧠 AI-Powered Repository Intelligence

Gain deep insights into large codebases through:

• Semantic repository search
• Repository-aware AI chat
• Knowledge graph generation
• Dependency analysis
• Architecture visualization
• Pull Request Intelligence
• Dead code detection
• Architecture drift analysis

💻 Modern Developer Experience

This release includes a redesigned developer experience powered by:

• FastAPI backend
• Astro frontend
• VS Code Extension
• Automated Docker builds
• GitHub Actions CI/CD

---

⚙️ Engineering Improvements

This release significantly improves platform stability and maintainability.

Highlights

• Resolved Pydantic compatibility for the latest MCP SDK
• Cross-platform CI stability across Windows and Linux
• Deterministic Ruff linting and formatting
• Improved dependency management
• Hardened automated test infrastructure
• Production-grade Docker release pipeline
• Automated GitHub Release workflow

---

📊 Release Quality

Repository Intelligence Agent v1.5.0 is backed by a comprehensive automated validation pipeline.

Metric	Status
Automated Tests	✅ 2505 Passing
CI/CD Pipeline	✅ Passing
Ruff Lint	✅ Passing
Ruff Formatting	✅ Passing
Docker Release	✅ Automated
Cross-Platform Testing	✅ Windows & Linux

---

🏗 Technology Stack

Backend

• Python 3.12
• FastAPI
• Pydantic v2
• ChromaDB
• Tree-sitter

Frontend

• Astro
• TypeScript

AI & Developer Tooling

• Model Context Protocol (MCP)
• GitHub Actions
• Docker
• Ruff
• Pytest

---

🎯 Project Vision

Repository Intelligence Agent is designed to make large software projects easier to understand.

Whether you're onboarding to a new repository, reviewing pull requests, exploring unfamiliar architectures, or building AI-assisted developer workflows, the platform provides intelligent context instead of requiring hours of manual code exploration.

---

📚 Documentation

Comprehensive documentation is available in the repository, including:

• Installation Guide
• Architecture Documentation
• MCP Integration
• API Reference
• Developer Guide
• Contributing Guide

---

🙌 Thank You

Thank you to everyone who contributed ideas, testing, reviews, bug reports, and improvements throughout the development of Repository Intelligence Agent v1.5.0.

Your feedback helped shape this first production-ready release.

---

⭐ Support the Project

If Repository Intelligence Agent helps you understand software faster or improves your development workflow, consider giving the repository a ⭐ Star on GitHub. Your support helps the project grow and reach more developers.

---

📦 Full Changelog

v3.0.0...v1.5.0

Repository Intelligence Platform v1.0.0

Repository Intelligence Platform v1.0.0

🎉 First stable production release of the Repository Intelligence Platform.

This release introduces a complete AI-powered repository intelligence platform with a FastAPI backend, Astro frontend, and a feature-rich VS Code extension for repository analysis, engineering insights, and developer productivity.

Highlights

Backend

• Repository Inspector
• Digital Twin
• Knowledge Graph
• Engineering Memory
• Graph RAG
• Reasoning Engine
• Structural Retrieval
• Advisor
• Execution Planner
• Workspace APIs
• Continuous Monitoring

VS Code Extension

• Repository Explorer
• Engineering Findings
• Advisor Dashboard
• Execution Planner
• Workspace Dashboard
• Repository Graph
• Repository Search
• Reading Path
• Timeline View
• CodeLens
• Hover Intelligence
• Code Actions
• Diagnostics
• Inline Decorations
• File Decorations
• Notification System

Quality & Validation

• ✅ 911 automated tests passed
  • 794 backend tests
  • 117 VS Code extension tests
• Production-ready documentation
• Comprehensive API reference
• Updated architecture documentation
• Production release tag: v1.0.0

Included Components

• FastAPI backend
• Astro frontend
• VS Code extension
• REST APIs
• Repository Intelligence Engine
• Engineering Analysis Pipeline
• Documentation and developer guides

Thank You

Thank you to everyone who contributed, reviewed, tested, and helped improve the project.