Monday, July 06, 2026

CloudNativePG 1.30.0 introduces a new DatabaseRole CRD for declarative, GitOps-style PostgreSQL role management and a Lease-based mechanism for safer primary election during failover. The release also includes notable security and operational improvements. Additionally, maintenance versions 1.29.2 and 1.28.4 are available, with 1.28.x reaching end-of-life on June 30, 2026.

• Use DatabaseRole CRD for declarative, GitOps-friendly role management
• Leverage Lease-based primitive for safer primary election failover
• Upgrade from 1.28.x before EOL on June 30, 2026
• Consider moving to 1.29 or 1.30 for latest security fixes

The pg_dbms_lock extension has reached version 2.0, a maintenance update that adds support for PostgreSQL 19. This release also aligns with the pg_background v2+ API, dropping support for older pg_background versions. The extension continues to emulate Oracle DBMS_LOCK behavior using PostgreSQL advisory locks, including lock modes, timeouts, and on-commit release settings.

• Update required for PostgreSQL 19 environments using pg_dbms_lock.
• Must upgrade pg_background to v2+ or break extension compatibility.
• Retains Oracle DBMS_LOCK emulation for exclusive/shared locks and timeouts.
• DBMS_LOCK.CONVERT() remains unsupported in this compatibility layer.

The credcheck extension for PostgreSQL has reached version 5.0, offering a comprehensive suite of credential checks triggered during user creation, password changes, and renames. This release allows administrators to define strict rules such as password expiration, reuse policies, brute-force protection via failure limits, and mandatory password changes upon first login. By integrating these checks directly into the authentication lifecycle, organizations can enforce stronger security postures without relying on external tools.

• Enforce password expiration, reuse policies, and minimum age requirements natively during user lifecycle events.
• Mitigate brute-force attacks by defining authentication failure limits and mandatory delays before banning users.
• Ensure compliance by forcing password changes on first login and issuing warnings before expiration.
• Integrate credential validation directly into PostgreSQL via the extension for centralized policy management.

Antony Pegg demonstrates provisioning a Retrieval-Augmented Generation (RAG) server on pgEdge Cloud using the API rather than the simplified UI. The guide provides concrete API calls to configure the service, emphasizing the need to manage API keys securely and avoid shell history exposure. The author uses a GURPS tabletop RPG dataset as the primary example for this setup.

• Use the Cloud API instead of the UI for reproducible, scriptable RAG server provisioning.
• Keep API keys out of shell history to prevent credential leakage during automated workflows.
• Ensure your dataset is ready and accessible before initiating the API configuration calls.
• Adapt the provided API structure to replace placeholder values with your specific environment configs.

AWS has increased the default runtime quota limits for Amazon Bedrock AgentCore to support larger-scale agent workloads. The update allows up to 5,000 active concurrent sessions in US East (N. Virginia) and US West (Oregon), while other regions support 2,500. Across all available regions, the service now handles 200 agent interactions per second and 25 new session creations per second by default.

• US East and West regions now support 5,000 active concurrent sessions by default.
• All other supported regions support 2,500 active concurrent sessions by default.
• Global throughput limit is 200 agent interactions per second across all regions.
• New session creation rate is capped at 25 per second in all regions.
• Higher default limits reduce the need to request quota increases for high-scale agents.

OpenAI has published a plugin enabling Claude Code users to invoke Codex directly within their existing workflow. The integration supports commands for standard reviews, adversarial challenges, and task delegation with background job management. Users with a ChatGPT subscription or OpenAI API key can install it via the marketplace to leverage Codex capabilities without leaving the Claude Code environment.

• Enables direct Codex invocation from within Claude Code workflows.
• Supports /codex:review and /codex:adversarial-review commands.
• Allows task delegation via /codex:transfer and /codex:rescue.
• Requires ChatGPT subscription or OpenAI API key for access.
• Consumes Codex usage limits associated with the user's account.

AWS has expanded Amazon Bedrock AgentCore availability to four new regions: Asia Pacific (Bangkok), Asia Pacific (Malaysia), Europe (Milan), and Europe (Spain). This platform enables engineers to build, connect, and optimize AI agents using any framework or model while enforcing security at the infrastructure layer. The expansion allows customers to run agents closer to end users, reducing latency while maintaining capabilities like identity management, tool connectivity, and observability.

• New regions support lower-latency agent deployment for users in Southeast Asia and Southern Europe.
• All core capabilities including runtime, policy management, and observability are available at launch.
• Security is enforced at the infrastructure layer, preventing agents from bypassing access controls.
• Engineers can connect agents to enterprise systems and optimize them continuously across these regions.

This GitHub project implements persistent file-based planning for AI coding agents using markdown files on disk. Version 3.0 introduces opt-in autonomous modes and a deterministic completion gate that prevents premature termination. It aims to solve context loss issues by maintaining task, findings, and progress state across agent resets.

• Skips context window limits by persisting planning state to disk files.
• New v3.0 completion gate ensures agents finish plans before stopping.
• Compatible with 60+ agents via the SKILL.md standard interface.
• Reduces hallucination risk by grounding decisions in explicit markdown plans.

Amazon SageMaker Unified Studio now supports infrastructure-as-code via Terraform, allowing teams to deploy domains using the terraform-aws-sagemaker-unified-studio module. This enables platform teams to integrate the unified AI development environment into existing CI/CD pipelines with version-controlled templates. The update ensures consistent workspace governance across development, staging, and production accounts.

• Deploy SageMaker Unified Studio domains using open-source Terraform modules for IaC consistency.
• Integrate AI workspace provisioning into existing infrastructure-as-code pipelines seamlessly.
• Maintain uniform access governance and catalog controls across multiple AWS accounts.

Amazon SageMaker HyperPod now provides visibility into the AMI versions deployed across training clusters and automates security patching. This eliminates the previous manual, reactive process that risked disrupting long-running foundation model jobs or altering critical bundled software like NVIDIA drivers. The update ensures cluster consistency and security without requiring operator intervention during sensitive workloads.

• Gain visibility into AMI versions across HyperPod clusters to detect drift quickly.
• Auto-patching applies security updates without disrupting multi-day training jobs.
• Reduces operational burden by removing manual patching workflows.
• Protects bundled software integrity, such as CUDA and NVIDIA drivers.

Amazon ECS now allows customization of deployment circuit breaker settings to control when a service deployment fails and triggers an automatic rollback. Users can define failure thresholds using either a fixed task count or a percentage of the desired task count. The feature also supports choosing between consecutive failure models to better align with specific application startup behaviors and failure tolerances.

• Customize rollback triggers using fixed task failure counts or percentages of desired tasks.
• Select consecutive failure counting models to match specific application startup patterns.
• Reduce manual intervention by automating rollbacks based on tailored failure thresholds.
• Improve deployment reliability by aligning circuit breaker logic with environment tolerance.

Amazon EKS now allows operators to revert to the previous Kubernetes minor version within seven days of an upgrade. The service performs automated readiness checks covering API compatibility, add-ons, and cluster health before executing the rollback. For EKS Auto Mode clusters, the system automatically manages worker node downgrades prior to reverting the control plane. Rollbacks can be triggered via the console, AWS CLI, or SDKs.

• Provides a 7-day safety window to revert Kubernetes minor version upgrades in EKS.
• Automated readiness checks validate API compatibility, version skew, and add-on status before rollback.
• EKS Auto Mode handles worker node downgrades automatically before control plane reversion.
• Initiate rollbacks using the EKS console, AWS CLI, or AWS SDKs.

AWS has extended ECS Express Mode to support custom task definitions, allowing users to apply the service's automated infrastructure management to existing application configurations. This update enables the reuse of task definitions from current CI/CD pipelines and IaC workflows without sacrificing the simplified deployment features like auto-scaling and load balancing. The change bridges the gap between rapid Express Mode deployments and complex operational requirements for containerized workloads.

• Custom task definitions are now valid inputs for ECS Express Mode deployments
• Retain existing CI/CD and IaC task definitions while using Express automation
• Gain automatic load balancing, networking, and scaling for custom apps
• Simplifies deployment of containerized web apps and APIs with advanced configs

Amazon RDS for Oracle now integrates with the Oracle Data Masking and Subsetting Pack, enabling secure data handling for non-production environments. The update allows users to configure masking policies directly through Oracle Enterprise Manager (OEM) and provides automation capabilities for routine data sanitization tasks. This feature helps organizations reduce compliance risks by ensuring sensitive information is obscured in lower-tier instances.

• Masking is managed via Oracle Enterprise Manager (OEM) interface for RDS instances.
• Supports both masking and subsetting to optimize storage and performance.
• Automation options available for scheduling regular data sanitization jobs.
• Helps meet regulatory requirements for PII and sensitive data protection.
• Simplifies setup compared to legacy manual masking scripts.

pg-cdc is a single-binary tool that reads Postgres Write Ahead Logs and writes typed, immutable Parquet files to AWS S3 in Iceberg format. It integrates with AWS Glue for metadata and enforces access control via Lake Formation tags, ensuring consumers never need database credentials. The architecture is strictly one-way, preventing any data from flowing back to the production database, which isolates analytics workloads from the source system.

• Single binary implementation with no JVM dependency reduces deployment overhead.
• One-way WAL stream guarantees analytics agents cannot modify production data.
• Lake Formation tags govern access, removing the need for DB credentials.
• Iceberg tables on S3 enable time-travel queries and immutable storage.

Meetily is an open-source, self-hosted AI meeting assistant built in Rust that processes audio entirely on local infrastructure. It leverages a 4x faster Parakeet/Whisper model for live transcription and speaker diarization, while using Ollama for summarization. The tool is designed for enterprises requiring data sovereignty, offering macOS and Windows support without cloud dependencies.

• Rust-based architecture ensures high performance for real-time audio processing tasks.
• Parakeet/Whisper integration delivers significantly faster live transcription speeds.
• Ollama handles summarization locally, maintaining strict data privacy and sovereignty.
• Fully self-hosted solution suitable for enterprise compliance and secure environments.

The asgeirtj/system_prompts_leaks repository catalogs extracted system instructions for major AI models including Anthropic's Claude series, OpenAI's GPT variants, and Google's Gemini. It highlights specific version changes, such as the transition from Claude Opus 4.8 to Fable 5, and includes prompts for developer tools like Cursor and VS Code. The collection serves as a reference for understanding the hidden rules governing these AI interactions.

• Provides raw system prompt text for top-tier models (Claude, GPT, Gemini).
• Tracks version diffs to identify instruction changes between model releases.
• Includes prompts for developer tools like Cursor and VS Code Copilot.
• Useful for prompt engineering and understanding model constraints.
• Aggregates data from multiple vendors into a single searchable repo.

The .NET team has released a curated repository of agent skills designed to assist AI coding agents with C# and .NET development. It includes plugins for language server integration, data access via Entity Framework, performance diagnostics, and MSBuild tasks, following the agentskills.io standard. The project provides a dashboard to track accuracy and efficiency trends for these contained plugins.

• Integrates LSP skills directly into AI coding agents for better C# context awareness.
• Offers specialized plugins for EF data access, diagnostics, and MSBuild operations.
• Follows the agentskills.io standard for interoperable AI agent capabilities.
• Provides a public dashboard to monitor plugin accuracy and efficiency metrics.