Microsoft RD-Agent: automated AI-driven R&D for data and model development
Microsoft has released RD-Agent, an open-source Python framework aimed at automating high-value R&D processes in AI, with a focus on data and model development. The project positions AI as the driver of data-driven AI workflows, targeting industrial productivity use cases. With 13,500 GitHub stars, it has attracted meaningful community interest, and a technical report is available.
Related guides (3)
Related events (8)
Microsoft agent-framework: open-source library for building and orchestrating AI agents
Microsoft has published an open-source framework on GitHub for building, orchestrating, and deploying AI agents and multi-agent workflows, with support for both Python and .NET. The repository has accumulated 11,061 stars. It represents Microsoft's entry into the agent harness tooling space alongside existing frameworks like LangChain and AutoGen.
Databricks AI Dev Kit: Toolkit for Coding Agents
Databricks Field Engineering has published an open-source Python toolkit on GitHub aimed at coding agents, accumulating 1,555 stars. The repository appears to provide utilities and scaffolding for building AI-powered development agents on the Databricks platform. It is a community/field engineering project rather than an official product release.
Microsoft RAMPART: pytest-native safety and security testing framework for agentic AI
Microsoft has released RAMPART, an open-source Python framework for safety and security testing of agentic AI applications, built natively on pytest. The repository is gaining traction on GitHub with 301 total stars and 77 new stars today. It targets the growing need for structured evaluation tooling specifically designed for AI agents rather than traditional software.
Microsoft Build: Seven in-house AI models, GitHub Copilot desktop agent manager, and Web IQ search API for agents
Microsoft announced seven new AI models trained from scratch (not distilled from OpenAI), including the flagship MAI-Thinking-1 reasoning model and MAI-Transcribe-1.5, plus a 'Frontier Tuning' reinforcement learning approach for enterprise workflow training. GitHub released a desktop Copilot app designed to manage multiple parallel AI agents with isolated git worktrees and bidirectional canvases. Microsoft also launched Web IQ, an agent-native Bing-powered grounding API already powering search in Copilot and ChatGPT, running 2.5x faster than alternatives with lower token costs. The roundup also covers Nous Research's Hermes Desktop cross-platform agent app, Alibaba's Qwen3.7-Plus multimodal model, and OpenAI's role-specific Codex plugins.
ARIS: Lightweight autonomous ML research agent using Markdown-only skills
ARIS (Auto-Research-In-Sleep) is an open-source Python project providing lightweight, framework-free Markdown-based skills for autonomous ML research workflows, including cross-model review loops, idea discovery, and experiment automation. It is designed to work with any LLM agent backend including Claude Code, Codex, or others. The project has accumulated 11,791 GitHub stars with notable daily traction (+106), suggesting meaningful community adoption.
pydantic/pydantic-ai: AI Agent Framework Trending on GitHub
pydantic-ai is an open-source AI agent framework built by the Pydantic team, applying Pydantic's data validation patterns to AI agent construction. The repository has accumulated 17,238 stars with modest daily momentum (+16 today). It represents a community-level signal of interest in structured, type-safe agent tooling in Python.
K-Dense-AI/scientific-agent-skills: Ready-to-Use Agent Skills Library for Research and Engineering
A Python repository providing a collection of pre-built agent skills targeting research, science, engineering, analysis, finance, and writing tasks. The project has accumulated 24,087 stars with a notable single-day gain of 762 stars, indicating significant community traction. No detailed technical documentation is available from the snippet, but the scope suggests a modular agent tooling library.
AutoResearchClaw: Fully Autonomous Self-Evolving Research Agent (Idea to Paper)
AutoResearchClaw is an open-source Python project from aiming-lab that claims to automate the full research pipeline from idea to paper, positioning itself as a fully autonomous and self-evolving research agent. The repository has accumulated 12,426 stars with 55 added today, indicating notable community traction. It represents a concrete implementation in the growing space of AI agents designed to conduct and write scientific research autonomously.