This week’s roundup covers new developments in agentic AI, streamlined team automation, and improved security. GitHub Copilot continues to evolve with context-sensitive agents and simplified integration through Model Context Protocol, while Azure delivers more automation for disaster recovery, AI-supported testing, and practical reliability resources. Updates for .NET 10 and emerging AI frameworks enhance workflow and testing, while recent security changes include identity recovery improvements and new fuzzing guidelines for open source code. These trends point towards more automated, interconnected, and secure development environments.

This Week’s Overview

GitHub Copilot

GitHub Copilot continues to play a larger role in development workflows, especially in Visual Studio Code and agent-driven automation. Current topics include using Copilot and Model Context Protocol (MCP) for agents that better understand workspace context, improved collaboration, and coding processes structured around clear specifications.

Context Engineering and Team Workflows in VS Code

Building on last week’s guidance for Copilot as a context-aware assistant in Visual Studio Code, this update covers detailed steps for setting up Copilot to behave consistently across teams. Previous discussions highlighted Agent Skills, structured prompts, and reusable instructions; now, the focus shifts to practical implementation using templates, custom guidance, and Copilot Plugins (MCPs) tailored for enterprise environments.

The article describes processes for encoding your project’s coding standards, architecture, and workflow guidance directly into your Copilot agent. It expands earlier updates on managing sessions and setup at the repository level, with hands-on templates and reference repositories that help teams achieve consistent Copilot behavior. These reusable practices support automation and peer review, giving teams a way to guide AI toward their established standards.

This content moves ahead from last week’s coverage of agent and workflow sharing—such as repository-wide settings and Skills.md—by providing tools that enforce norms and support larger collaborative engineering teams.

Agentic AI, MCP Integration, and Spec-Driven Development

After recent deep-dives into Copilot’s agent-driven modes and use of MCP, the latest resources demonstrate more cross-agent project work and improved model integration. Last week covered Agent Mode, the Cloud Agent, and support for BYOK (Bring Your Own Key) automation in VS Code; this week examines the new features in Copilot Agent Mode and stronger IDE integration.

The article highlights how automation such as repository scanning, code editing, and pull request support now benefits from better MCP integration. Copilot coordinates multiple agent types, including Anthropic, OpenAI, and Google models, all managed under one subscription—a natural extension of previous collaborative updates.

Agent HQ, introduced at GitHub Universe, supports community agent sharing and mult-agent workflows. The linked article explains how Spec Kit—mentioned previously as the base for specification-driven development—empowers robust, repeatable, and maintainable automation across teams.

Altogether, these articles show how Copilot is moving from simple code completion towards a platform supporting customizable, team-managed agents. The introduction of the MCP Registry and more case studies demonstrates how these new approaches are being put to use. Teams can now move beyond suggestions to structured, collaborative AI tools suited to company standards.

AI

This AI section highlights continued momentum in deploying Microsoft Agent Framework, evolving agent skills for enterprise software reliability, and new context engineering practices for AI-powered SRE tools. The role of AI in language trends for developers is also examined.

Context Engineering in Azure SRE Agent Development

Following last week’s overview of Microsoft Agent Framework for Azure SRE Agent automation, this technical review shares lessons in designing operational context for reliable AI agents. The Azure SRE Agent team describes methods for maintaining context boundaries, keeping track of state, and embedding guardrails for predictability—building on last week’s topics around production readiness and orchestration.

The article shifts the focus from diagrammatic design to real-world deployment. It explains hands-on guidance for structuring agent access, monitoring state, and ensuring operations within managed guardrails, situating these as practical steps to make enterprise agents dependable and supportable. The techniques serve organizations scaling agent use for critical workloads.

AI’s Impact on Programming Language Selection

This article offers a follow-up to recent discussions about workflows and language choice in AI development, with a focus on MCP and language-model compatibility. A new analysis from GitHub discusses how AI tooling is encouraging the use of statically typed languages, such as TypeScript, to improve reliability—a trend supported by recent Octoverse data showing increased adoption of those languages.

The video details the pattern: AI models are most effective with codebases that include clear type information, pushing developers to adopt languages that maximize AI’s benefit. It encourages teams to consider how language selection impacts tooling support and outcomes, especially with AI growing as a default part of the software development process.

Azure

Azure news this week focuses on new resources for disaster recovery, data migration, cloud backup, and automated test creation. Step-by-step guides support business continuity, resilience, and innovation in Azure environments.

Guides for Resilience, Migration, and Data Protection on Azure

Developers have three new guides for improving business continuity, streamlining analytics migrations, and automating data backup in Azure. The disaster recovery tutorial covers Azure Databricks and Microsoft Fabric, with instructions for using Terraform and CI/CD to replicate critical data services like ADLS, Key Vault, and SQL. Those working with Databricks get a Python solution for automating Databricks failover, while Power BI and Fabric users benefit from geo-redundancy and cross-region features aligned with SLAs.

Expanding on last week’s Azure Databricks best practices around security and data management, these guides extend to backup and restoration workflows and support compliance standards. Power BI and Fabric’s redundancy features build on earlier themes of reliability, automation, and ease of recovery.

For teams modernizing analytics, there is a recipe for migrating from Tableau to Power BI on Microsoft Fabric. The step-by-step migration focuses on a “semantic layer first” design—meaning that business logic is centralized for less duplication and easier governance. The guide covers everything from asset mapping and workspace setup to AI integration with Power BI Copilot and ML notebooks, and includes best practices on security, naming, and model design to help ensure long-term analytics maintainability.

An additional resource describes automating backup and restore for Azure Cosmos DB and Databricks, with support for self-service restores using Apache Spark. This supports scenarios like point-in-time recovery, test environment deployment, or compliance reporting. Restores can be executed before deployment or to quickly roll back production, and pipelined with CI/CD for easy workflows. Documentation and example cases make these features accessible for different organizations.

All three guides expand last week’s focus on cloud data security and automation, now addressing everything from scheduled failover to migration playbooks and one-click database recovery.

AI-Assisted Load Testing and Performance Workflows

A new set of AI-enabled tools for Azure App Testing allows teams to create load tests automatically. By using an Edge or Chrome browser extension, user sessions are captured and converted into JMeter scripts with the help of AI, including labeling, parameterization, and dynamic correlation. Scripts are fully editable for custom needs and can be used directly with Azure Load Testing for realistic test runs at scale. Version 4.0 focuses on minimizing manual work and supporting realistic performance engineering through feedback-driven improvement.

This extends recent updates introducing AI-powered load generators in Azure. The new tools further support automated performance scenarios and make integration into existing DevOps workflows easier.

Other Azure News

The Azure SRE Agent Community Hub is now live, providing SREs and cloud app developers with a central place to trade strategies, problem-solving tips, and updates. It includes forums, a dedicated blog, and the latest product information, all designed to foster knowledge sharing for those developing reliability automation on Azure.

This resource addresses recent requests for spaces to exchange practical knowledge and build collective experience with Azure reliability tools. It reflects two recent weeks of emphasis on site reliability engineering and agent-based automation.

Coding

This Coding section features updates for .NET, highlighting advances in .NET 10, opportunities for skill-building, and new resources in AI integration and developer education.

.NET 10 Technical Developments and AI Integration

The latest roundup on the official .NET blog recaps the .NET 10 LTS release and key additions to the runtime, SDK, and C#. Recent features include faster garbage collection, new benchmarking tools, and improved solution management with updated SLNX and CLI capabilities. .NET 10 also supports AI agent development with the Microsoft Agent Framework, upgrades to the MCP server, and instant access to AI via NuGet packages, helping teams use generative AI across both existing and new solutions.

This post continues threads from the past week’s MCP coverage and agent automation, now available through official templates and NuGet for easier use. The Agent Framework and server reduce barriers for customizing enterprise AI agents. Copilot integration, now through enhanced LTS toolchains, connects .NET’s automation approach with that of Visual Studio Code. Articles compare Copilot’s new Ask and Agent Modes, outlining how they fit into .NET development.

Additionally, .NET Aspire 9.3 brings more support for cloud architectures and improved onboarding. The post links to deep dives, community highlights from .NET Conf 2025, security news, and official lifecycle policies, helping developers stay current with migrations and updates.

.NET Developer Education and Video Resources

Jon Galloway’s annual roundup curates the year’s top videos and live streams for .NET developers. The featured content spans demos of .NET 10 and Visual Studio 2026, Clean Architecture, performance tuning, modular monolith patterns, advanced C# 14 topics, and Blazor with AI integration (including Python and MCP demos). Tutorials cover a range of workflows, including Hands-on with Blazor AI templates, and track .NET’s continued intersections with modern AI.

This complements ongoing themes of open source development, skill advancement, and innovation through community events. Demo sessions, podcasts, and experimental MCP showcases follow up on discussions from the latest podcasts and technical experimentation.

Community Standup events and “Deep .NET” livestreams highlight interactive and transparent development processes. Additional topics include deploying AI locally, using Akka.NET for high-throughput data, and optimizing SQL. All content is freely available on YouTube and Microsoft Learn to help developers build current skills and adopt new .NET concepts.

DevOps

This week in DevOps, Azure added tools for AI-guided debugging and repair, while GitHub made it more affordable to run hosted CI/CD runners. Both changes support smoother operations and accessibility in cloud engineering.

AI-Powered Debugging and Remediation with Azure SRE Agent

Extending coverage of SRE Agent automation and incident workflows powered by MCP, the Azure SRE Agent now includes developer-focused support for debugging and automated remediation. This week’s workflow goes beyond detection, connecting infrastructure monitoring and health analysis to agent-driven fixes.

The latest advances combine VS Code Copilot and Claude Opus for more effective debugging. The GitHub Coding Agent structures pull requests for all types of changes, unifying app, infrastructure, and SQL repairs. This parallels recent articles arguing for source-controlled change automation as a safer choice than ad-hoc fixes in cloud portals.

The automation tools benefit from improvements in orchestration and permission models introduced earlier, and reinforce a shift toward open source agent engagement. Developers are encouraged to contribute and test agent features as the model for custom MCP adapters and integrated alerts expands.

Lower Costs for GitHub-Hosted CI/CD Runners

GitHub Actions pricing for hosted runners has been reduced as of December 16, 2025, supporting the continued push for easy and affordable CI/CD pipelines. Private repository users can now automate builds and releases at a lower cost, with public repositories retaining free runner access. New calculators help teams predict costs and plan for automation at different scales.

This change follows last week’s news about workflow reliability and aligns with the ongoing goal of making GitHub Actions practical and accessible for development teams. Cost savings should help further automate builds and expand testing in private repos, fitting the recurring story of broader CI/CD adoption.

Security

This week’s security updates include a research review on continuous fuzzing in open source and new features in Microsoft Entra ID for more resilient cloud identity management and user recovery.

Fuzzing and Vulnerability Discovery in Open Source Projects

GitHub Security Lab recently published findings on why some security bugs persist in open source projects even after extended fuzzing with OSS-Fuzz. Case studies point out that incomplete coverage—especially around encoding logic and external dependencies—lets vulnerabilities survive. Issues sometimes escape detection because fuzzers are not running long enough or do not generate large enough inputs. The article explains advanced options like AFL++ branch coverage, N-Gram, value-based fuzzing, and the addition of manual reviews or static analysis for better detection. It lays out five steps for closing test coverage gaps and suggests using Fuzzing 101, workflow reviews, and layered assurance.

These recommendations add to last week’s supply chain security topics, arguing that effective security testing requires a mix of automation and human review. Open source maintainers are reminded that complex bugs sometimes need multiple validation steps to be found and fixed, reinforcing the ongoing emphasis on persistent, multifaceted vulnerability discovery.

Improvements to Account Recovery in Microsoft Entra ID

Microsoft Entra ID (formerly Azure Active Directory) now offers expanded options for user account recovery. In addition to password reset and SMS, blocked users can restore access by submitting a government ID for third-party verification (through services like AU10TIX). This approach is designed to resist phishing and credential theft, lessening social engineering risk and SIM swap attacks. Setup is done through a portal integration and Azure API configuration, giving IT administrators a way to enable or restrict the feature.

Privacy, regulatory policies, and provider stability are addressed, and step-by-step instructions with demo videos are available for deployment. Improved account recovery helps reduce support tickets and lets users regain access without as much IT involvement—a boost for cloud identity stability.

This update fits with previous work to raise the bar for identity security, following MFA, strong authentication, and trusted publishing as mentioned previously.