AI Briefing#open-source
The latest AI and developer news about #open-source, with the original source and a short summary.
The latest AI and developer news about #open-source, with the original source and a short summary.
AgentRail, an open-source task lifecycle layer that supports task management and CI/CD integration for coding agents, has been released.
Hugging Face has released ml-intern, a tool for agents that controls ML libraries using local models.
The PyTorch library for Hessian/GGN analysis has been rewritten as v1.0.0a1.
Raindrop Workshop, an open-source local tool that lets you trace and debug AI agent execution token by token, has launched.
opendesk, an open-source tool that lets AI agents control other computers over WiFi, has been released.
opendesk, an open-source tool that lets AI agents remotely control other computers over WiFi, has launched.
An open-source Mac app has launched that monitors the work of Claude Code and Codex agents in real time.
Claude Flow has been rebranded as Ruflo and released as a platform with over 100 multi-agent capabilities for Claude Code.
Audrey 1.0, a local-first memory and control layer for Claude Code-style agents, has been released.
TrustClaw, an open-source AI agent that can be easily deployed on Vercel, has been released.
Resemble AI has released DramaBox, a voice synthesis model based on LTX 2.3.
Cline has released the Cline SDK, a plugin-based open-source agent runtime.
A summary of 10 open-source roguelikes that survived through community forks and maintenance.
You can integrate Claude Code with Ollama models via the open-source LLM router Manifest to reduce costs.
An open-source MCP tool has been released that lets AI agents directly control the desktop by leveraging the OS's accessibility layer.
PullMD v2.4.1 has been released, natively supporting claude.ai web custom connectors.
llama.cpp's server and webui now support the continue generation feature for reasoning models.
Open Browser Use connects local AI agents to a real Chrome profile.
Needle, a distillation of Gemini tool calling into a 26-million-parameter model, has been released.
Open Computer Use has released local desktop automation as an MCP service.
A terminal coding agent based on DeepSeek V4 has been released.
Luce DFlash/PFlash boosts LLM inference performance by up to more than 3x on the AMD Strix Halo iGPU.
Needle has open-sourced a Gemini-based tool-calling model with 26 million parameters.
Hugging Face has released 'physics-intern,' a multi-agent framework for theoretical physics research.