Browse Security News (264)

Richard Lander explains how NuGet packages can carry version-matched agent skills alongside library binaries. He demonstrates tools for discovering and copying those skills into repositories, compares package delivery with Git-based approaches, and discusses model evaluation, token costs, and the security implications of agents consuming package documentation.

tompeplow demonstrates how OneLake supports agent-driven data engineering through hosted MCP servers, endpoint discovery, and shortcuts. Using GitHub Copilot, the examples explore mirrored databases, document permissions, combine external data in a Lakehouse, and run SQL and Spark analytics while working within existing access permissions.

Microsoft Security Research and Jiacen Xu examine a fake-download campaign that bundles proxyware, distributed-computing software, and unwanted utilities through trojanized installers. They explain its staging and callback patterns, scheduled-task persistence, and Microsoft Defender hunting queries, with guidance for correlating evidence across an entire installer session rather than individual payload attempts.

Yun Jung Choi explains how six specialized Azure plugins extend AI-agent workflows for AKS troubleshooting, cost analysis, Advisor recommendations, graph queries, Azure Local, and Foundry IQ knowledge bases. The guide includes example prompts, installation commands for supported clients, and approval requirements for resource changes and disruptive operations.

Allison outlines CodeQL 2.27.3 updates for GitHub code scanning, including Python 3.15 support, JSON data extensions, and improved analysis across languages. The release also adds an opt-in threat model for LLM-generated data and introduces GitHub Actions control flow changes that may require updates to custom queries.

Allison announces that Dependabot now supports Python 3.15, allowing projects to upgrade while continuing to receive dependency updates. Support is available to all github.com users and will also be included in GitHub Enterprise Server 3.24.

Allison explains how GitHub Desktop 3.6.10 and later use short-lived, automatically refreshing authentication tokens for eligible accounts. The gradual rollout reduces leaked-token exposure without manual renewal for Git operations or API requests. Existing sessions remain signed in, but existing tokens are not automatically migrated, and GitHub Enterprise Server is excluded.

sbaynes shares how Microsoft's Copilot Cowork team used coding agents, written specifications, automated reviews, and human checkpoints to build and migrate a live service. The article explains practical controls for parallel development, security boundaries, trustworthy tests, and a Python-to-Rust migration, including measured gains and remaining limitations.

Microsoft Security Research, Yair Tsarfaty and Or Begam analyze which vulnerabilities attackers actually exploit and why system-takeover flaws deserve priority. Their findings connect exploitation patterns and patching delays with practical use of Microsoft Defender exposure scores and attack path analysis to reduce risk beyond patching alone.

Allison explains how GitHub’s code security and secret risk assessments now distinguish findings in protected and unprotected repositories. The update helps organization admins and security managers identify gaps in security coverage, with availability in GitHub Enterprise Cloud and GitHub Team and planned secret assessment support in Enterprise Server 3.24.

Allison outlines GitHub Actions’ upcoming default block on pull_request_target events in public repositories, with enforcement moved to December 2, 2026. Two November brownouts will help maintainers identify affected workflows. The update explains how to review blocked runs and explicitly permit workflows while accounting for risks to repository secrets.

Microsoft Security Research analyzes a malicious Claude Code-branded extension distributed through Cursor’s marketplace, detailing its attempted SSH backdoor and cross-platform persistence. The article provides Microsoft Defender detections, hunting queries, and remediation steps, explaining why removing the extension alone does not eliminate unauthorized keys and startup artifacts.

Microsoft Security Research, Sagar Patil, Jesse Birch and Tal Tzhori examine two intrusions that escalated from help desk impersonation to AD CS certificate abuse. Their analysis explains ESC1 and ESC8 attack paths, contrasting containment outcomes, and practical Microsoft Defender hunting queries and hardening measures for certificate-backed identity compromise.

spurthikommajos explains how Microsoft Fabric Eventstream’s preview support for workspace-level private connectivity lets custom applications publish events without public internet access. The article covers Azure private endpoints, workspace-scoped network restrictions, setup prerequisites, and supported analytics destinations, including important capacity requirements and ingestion limitations.

yaelbiss explains how Microsoft Fabric’s preview Allow workspace settings editing policy restricts sensitive encryption and networking changes without removing workspace administrator responsibilities. The article covers security-group-based rules, development and production boundaries, audit logging, and allow-list behavior that administrators should test before activating policies across a tenant.

amitch and Shehab Amin explain how Sail connects Spark-compatible workloads to Microsoft OneLake without copying data. They demonstrate lakehouse transformations, Python-based embedding generation, and scoped access through LakeSail Cloud Sessions, while outlining compatibility checks and limitations of the reported performance and cost comparisons.

Allison details three additional IP ranges for GitHub webhook deliveries, now included in the Meta API. Teams using static network allowlists need to add the prefixes to avoid rejecting deliveries. Allowlists kept current through the Meta API require no changes, and the additions apply across all github.com plans.

Salim Chawro introduces Hardware Audit Logging for Azure Cloud HSM, which records supported management operations directly from the hardware. Signed audit data and certificates enable independent verification, while Azure Monitor Log Analytics delivery supports investigations and monitoring. The feature is rolling out across public regions and is automatically enabled on new resources where available.

Microsoft Threat Intelligence examines how North Korean threat actors use AI to accelerate reconnaissance, identity deception, malware development, and operational scaling. The report distinguishes observed activity from suspected AI use, explains why operations remain human-directed, and provides Microsoft Defender detections and practical guidance for protecting identities, endpoints, and software repositories.

nareshvenkat introduces Project Osmos, a data engineering agent for Microsoft Fabric that investigates datasets, executes PySpark, and validates results. Persistent, shared tasks retain project context while teams define operating boundaries, review evidence, and guide changes. Tasks can be initiated through GitHub Copilot CLI, Claude Code, or Codex.

LeeBenjamin introduces the generally available Govern experience in Microsoft Fabric’s OneLake catalog and the new Govern Skills collection. Together, they provide centralized estate insights, policy controls, and AI-assisted governance checks, with evidence-based recommendations and confirmation before sensitive changes through tools such as GitHub Copilot.

Lynn Dai introduces Trusted Publishing for Visual Studio Marketplace, enabling extension releases from GitHub Actions and Azure Pipelines without long-lived personal access tokens. The article explains federated identity setup, GitHub policy activation, Azure Pipelines service connections, and how to migrate existing publishing workflows while verifying access before removing stored secrets.

GitHub Staff examines the 2026 Octoverse findings on repository activity, agent-started pull requests, and changing model choices. The report connects growing use of coding agents with shifts in review workflows, GitHub Actions usage, and language participation, while explaining the telemetry’s limitations and announcing general availability of GitHub Copilot Fusion.

Allison introduces the GitHub security extension, now in public preview for the GitHub Copilot app. Developers can browse code scanning alerts, ask Copilot to explain or fix them locally, and review changes before opening pull requests. The announcement also covers permissions, AI credit usage, and supported GitHub environments.

Allison announces general availability of GitHub Agentic Workflows, which compiles natural-language Markdown into GitHub Actions workflows powered by coding agents. The release adds runner support, model routing, and stronger isolation, with read-only defaults and validated outputs. The default GitHub Copilot CLI engine requires a Copilot license.

Allison introduces Dependabot reachability analysis, now in public preview for JavaScript and TypeScript on GitHub Enterprise Cloud. Using CodeQL and managed GitHub Actions workflows, it identifies call paths to vulnerable dependency functions and adds verdicts to alerts, helping teams prioritize remediation without automatically dismissing vulnerabilities.

Sandeep Sen summarizes October's Azure SDK releases, including authentication changes in Azure Core for .NET, managed identity improvements for Python and Java, and the first stable Cosmos DB libraries for Rust. The update also lists new stable management libraries and initial previews for AI and provisioning libraries.

Allison announces public preview support for server-to-server authentication in GitHub Copilot. GitHub Apps can use short-lived installation tokens to make SDK requests on behalf of an organization without user credentials. Usage is billed to the organization, with its model policies, budgets, spending controls, and app access controls applied.

Allison announces general availability of assisted permissions in GitHub Copilot CLI, GitHub Copilot app, and Visual Studio Code. A language model evaluates command, file-path, and URL requests, automatically allowing those classified as clearly safe. Enterprise administrators can control available permission modes and set defaults for new sessions.

sbaynes describes how Microsoft combines AI agents, testing, static analysis, and formal verification to transform large codebases. The article covers Windows toolchain migration, C/C++ modernization into Rust, and specification-driven validation, explaining how small expert teams review shared knowledge and system properties rather than every generated change.

Sverre Brandsberg Dahl describes how SLB agile reservoir modeling connects with Azure Data Manager for Energy to run reservoir ensembles at cloud scale. The architecture combines parallel simulation, shared Azure Files storage, OSDU-aligned data services, and Microsoft Entra ID access controls to keep modeling workflows connected to governed corporate data.

Cynthia Zanoni and Pooja Yarabothu explain how Microsoft’s PostgreSQL extension brings database exploration, SQL editing, and troubleshooting into VS Code. They show how GitHub Copilot can help interpret schemas and query plans, while emphasizing result verification and careful review of performance recommendations before making changes.

Josef Sin explains what Azure DevOps users should check following the PHP2Shell disclosures. Azure DevOps Services and Server do not use PHP to handle customer requests, but pipeline workloads, agent hosts, containers, and deployed applications may require runtime patches. The guidance distinguishes Microsoft-managed updates from fixes customers must apply themselves.

Rahul Bhandari (MSFT) announces out-of-band security updates for .NET 8, 9, and 10, plus .NET 11 RC3. The release fixes Brotli decompression and Linux tarball ownership vulnerabilities, completes two earlier fixes, and explains how to update framework-dependent applications, self-contained deployments, containers, and source-build packages.

Allison explains a GitHub Enterprise Managed Users (EMU) incident where SCIM deprovisioning incorrectly deleted accounts instead of suspending and obfuscating them, how to confirm impact via audit logs, and what GitHub changed to restore records and prevent recurrence.

Microsoft Security Research, Mahesh Mandava, Srinivasan Govindarajan, Sagar Patil and Suriyaraj Natarajan break down the LiteLLM 1.107.0 supply-chain compromise, showing how Python startup hooks enabled credential theft without importing LiteLLM, what data was targeted, and how to investigate and recover using Microsoft Defender detections and hunting queries.

Allison announces new GitHub public monitoring capabilities for handling leaked secrets at enterprise scale, including alert triage workflows, continuous token validity checks, and integrations via a REST API and enterprise webhooks for incident response pipelines.

Allison announces general availability of enterprise-level installations for GitHub Apps, expanding API and permission coverage so enterprises can migrate more automation away from classic PATs and manage governance, security, and admin settings through GitHub Apps.

Allison summarizes the September 2026 update for GitHub Copilot in Visual Studio 2026, including PR exploration via the Git agent, preview BYOM support (with a breaking change to the new Agent harness), and a new flow to fix NuGet Audit vulnerabilities directly from the Error List.

Allison announces new GitHub Actions blob storage hosts that are now included in the GitHub meta API, with guidance for organizations that use network allowlists to avoid interruptions to job summaries and logs.