Ingress2Gateway 1.0 has been released to aid migration from Ingress-NGINX to Gateway API before its retirement in March 2026. The tool translates Ingress resources to Gateway API and highlights untranslatable configurations. The release features enhanced annotation support and thorough testing for reliable migration.

NanoClaw integrates with Docker Sandboxes to enhance AI agent security through strong isolation and transparency. This collaboration focuses on enabling secure and autonomous operations for AI agents within enterprise environments.

Microsoft's Project Silica encodes data in borosilicate glass using femtosecond lasers, offering long-term storage for up to 10,000 years. This method overcomes traditional storage limitations and is cost-effective, though write speed remains a challenge. The research phase is complete, but no product release has been announced.

California's Digital Age Assurance Act mandates operating systems to share users' age data with app developers via a real-time API by 2027. The law faces criticism for depending on self-reported ages, potentially affecting its efficacy.

NanoClaw is a lightweight open-source personal AI agent that runs locally and connects to apps like WhatsApp and Telegram. Built with only ~3,900 lines of code across 15 files, it uses container isolation to securely run agents and aims to offer a simpler, fully auditable alternative to large frameworks like OpenClaw.

OpenAI released GPT-5.4, its new flagship model for ChatGPT, the API, and Codex. It improves reasoning, coding, and agent workflows, introduces native computer-use capabilities, supports up to 1M tokens of context, and adds tool search to make large tool ecosystems more efficient. The model is more accurate, more token-efficient, and better at real professional tasks like coding, spreadsheets, documents, and web research. A higher-performance GPT-5.4 Pro version is also available for complex workloads.

FastMCP 3.0 is now generally available. It keeps the @mcp.tool() API but rebuilds the internals around components + providers + transforms, adds a CLI, and ships production features like component versioning, per-component auth + OAuth additions, OpenTelemetry tracing, background tasks, pagination, tool timeouts, and hot reload. The project moved from jlowin/fastmcp to PrefectHQ/fastmcp on GitHub, and upgrading is supported via dedicated guides for FastMCP 2 and MCP SDK users.

Peter Steinberger, creator of OpenClaw, is joining OpenAI to work on bringing AI agents to a broader audience, while OpenClaw will move to an independent open-source foundation and continue development outside OpenAI’s direct control.

Sipeed has released PicoClaw an OpenClaw micro alternative that uses 99% less memory than . , an open-source AI assistant written in Go that runs in under 10MB of RAM and boots in about one second. Designed for low-cost Linux boards starting around $10, it supports multiple LLM providers, chat platform integrations, and automation workflows. The project is MIT-licensed and available on GitHub.

GitHub has released the Copilot SDK in technical preview, allowing developers to embed Copilot’s agentic execution loop into their own applications. The SDK supports multiple AI models, real-time streaming, and languages like Python, TypeScript, Go, and .NET, but currently requires a Copilot subscription and is intended for development and testing rather than production use.