Live
GPU‑Accelerated Inference on AWS Lambda via Managed InstancesClaude Code’s “Don’t Ask Again” local file leaks credentials to public reposPost‑training RL with PivotRL lifts Nemotron 3 Ultra coding benchmarkIntegrating AI Coding Agents into Development Workflows: Protocols, Practices, and Security ImplicationsAWS Quick’s new Meta ads AI connector: what engineers need to knowAI router data leakage forces engineers to rethink prompt pipelinesRust’s Agent Client Protocol simplifies AI‑agent integration for production workloadsUsing Gaussian Sharpening with Cloudflare Images: Practical Guidance for EngineersGPU‑Accelerated Inference on AWS Lambda via Managed InstancesClaude Code’s “Don’t Ask Again” local file leaks credentials to public reposPost‑training RL with PivotRL lifts Nemotron 3 Ultra coding benchmarkIntegrating AI Coding Agents into Development Workflows: Protocols, Practices, and Security ImplicationsAWS Quick’s new Meta ads AI connector: what engineers need to knowAI router data leakage forces engineers to rethink prompt pipelinesRust’s Agent Client Protocol simplifies AI‑agent integration for production workloadsUsing Gaussian Sharpening with Cloudflare Images: Practical Guidance for Engineers
GitHub

GitHub Enterprise Server 3.23: New Project Filters, On-Prem AI Secret Scanning, and Token Expiration Controls

AI SummaryPowered by AI

GitHub Enterprise Server 3.23 adds logical‑operator filters in Projects, on‑prem AI‑driven secret detection, expanded malicious‑package advisory coverage, a three‑day Dependabot cooldown, optional eight‑hour OAuth token expiration, and the ability to place the Redis primary on a dedicated node. These changes affect how AI, cloud, DevOps, and security teams build, automate, and protect code and infrastructure on GHES.

GitHub Enterprise Server 3.23 introduces a set of features that touch project management, supply‑chain security, authentication, and infrastructure isolation. For engineers who automate pipelines, protect secrets, or run GHES in restricted networks, the new capabilities change how you configure and operate the platform.

Enhanced Project Querying

Projects now support combined filter expressions using OR, explicit AND, and parentheses for grouping. For example, (status:Done OR status:Todo) is:issue returns issues matching either status. This gives AI engineers and SREs finer control when building dashboards or automated triage bots that rely on precise issue views.

On‑Prem AI Secret Detection

Enterprises that enable GitHub Secret Protection can run AI‑based secret detection entirely within their network. A fine‑tuned model executes on the appliance’s CPU—no GPU, hosted LLM, or outbound connection is required. The feature is in public preview and works in air‑gapped environments, offering a way to scan for unstructured secrets without exposing code to external services.

Broader Malicious Package Advisory and Dependabot Timing

The advisory database now pulls malicious‑package data from OpenSSF for PyPI, Maven, RubyGems, NuGet, Go, Rust, and Composer, extending coverage beyond npm. Dependabot also enforces a three‑day cooldown before opening a version‑update pull request, giving maintainers time to identify compromised releases. Both changes aim to reduce supply‑chain risk for DevOps pipelines.

OAuth Token Lifetime Controls

OAuth app developers can opt‑in to eight‑hour access tokens with refresh tokens. New apps default to this expiration model; existing apps must enable it explicitly. Requests include scope=offline_access to obtain refresh tokens. This shift requires client updates to store and use refresh tokens, impacting authentication flows and token‑rotation policies.

Redis Primary Isolation

Administrators can now move the Redis primary service to a node dedicated to Redis, separating it from other services such as MySQL. The dedicated node’s resources and data volume are used exclusively by Redis, without attaching a new disk or relying on an external Redis service. This isolation can improve performance predictability and simplify capacity planning.

Related CloudNinjas coverage: DevOps.

What This Means For Practitioners

Evaluate the release candidate in a non‑production cluster to verify filter syntax, AI secret detection performance on your CPU, and the behavior of the Dependabot cooldown. Review OAuth app configurations and plan migration to refresh‑token flows before the default expiration becomes mandatory. If you rely on Redis, test the dedicated‑node procedure and monitor resource usage. Finally, provide feedback to GitHub on any preview‑only behavior that could affect your production rollout.

Originally published atGitHub Changelog