Open source news — new project releases, community governance, licensing updates, and open-source AI models including Llama, Mistral, and others.
ClawBattle, a new open-source benchmark that tests multimodal LLMs on code golfing tasks, which require both visual and textual understanding. It claims the benchmark avoids data contamination by using confidential, top-…
At Google I/O 2026, the company announced that its open-source Android app, AI Edge Gallery, now supports agent functionality and MCP (Model Context Protocol) connections, allowing large language models to run entirely o…
Google is discontinuing its open-source Gemini CLI tool and replacing it with a closed-source, proprietary version. Most users will lose access to the original tool on June 18, unless they have enterprise credentials or …
The Model Context Protocol (MCP) is an open standard published by Anthropic that enables AI assistants to securely discover, call, and receive results from structured tools, bridging the gap between AI and CRM systems. T…
ArcBrush is a free, node-based 2D image editor for Windows, macOS, and Linux that allows users to build non-destructive, automated pipelines for generating multiple image variants. By wiring together 79 different nodes, …
Release of mcp-probe v1.0.0, a CI-ready readiness gate for MCP servers that goes beyond basic startup checks to validate whether tools actually work in agent workflows. The tool can test real tool calls with configurable…
A developer created ValoVault, a third-party tool that automatically swaps skin loadouts in Valorant based on which agent a player selects, addressing a long-standing community request that Riot Games has declined to imp…
General availability of GKE Agent Sandbox, an open-source, Kubernetes-based execution environment designed for secure and scalable AI agent workloads, featuring capabilities like pod snapshots for idle compute reduction …
Google has introduced Agent Executor, an open-source runtime standard designed to reliably manage long-running, distributed agent workflows in production. The system addresses key operational challenges by providing dura…
AI coding agents fail senior engineers because they skip critical architectural planning and long-term system design, optimizing only for immediate code generation. To address this, the author built Kavro, an open-source…
The Gemma 4 26B model was the first local AI to achieve a perfect score on the HumanEval benchmark, including solving the notoriously difficult problem 145. This problem requires sorting integers by the sum of their digi…
Creation of Openbench AI, a desktop chat application designed to run local large language models (LLMs) offline via Ollama, eliminating the need for cloud services or complex setup. The app allows users to run multiple m…
Stable Audio 3 is a family of fast latent diffusion models (small, medium, large) designed for variable-length audio generation and editing, capable of producing several minutes of sound. The models use a novel semantic-…
A new Claude skill called "spec-driven-development" that creates standardized specification files (requirements.md, design.md, tasks.md) and matching configuration files for multiple AI coding tools (Claude Code, Cursor,…
Stability AI has released Stability Audio 3.0, a new family of audio models that includes a large model capable of generating professional-grade music compositions over six minutes long. The company is making the small a…
The Metabase 61 update introduces granular AI governance controls, allowing administrators to manage AI feature access per user group and set token limits for internal and embedded analytics. New features include an MCP …
Release of datasette-agent-charts version 0.1a1, which introduces several improvements including enhanced color schemes for bar and waffle charts, interactive tooltips, and a permission check for SQL execution. It also f…
"A Fact A Day," an autonomous daily educational podcast that summarizes scientific and technological breakthroughs in about 60 seconds. The entire production process, from researching sources like MIT Tech Review to scri…
The author conducted a six-week benchmark comparing ten agentic AI frameworks—including LangGraph, CrewAI, and AutoGen—by running identical tasks across five evaluation dimensions: setup time, tool integration complexity…
Snyk's agent-scan tool analyzes MCP servers by actually executing them from the user's configuration files, retrieving tool descriptions via the MCP protocol, and sending that data to a third-party API for analysis. This…