AI Dev Stack

Sign in Subscribe

feature

Cline SDK: The Modular Agent Runtime Powering 7M Developers

Cline SDK: The Modular Agent Runtime Powering 7M Developers

Cline rebuilt its agent runtime from scratch as an open-source SDK. The new harness scores 74.2% on Terminal Benchmark, ships with plugins and multi-agent teams built in, and powers agents across VS Code, JetBrains, and CLI.

Building a Roguelike from Your Codebase with GitHub Copilot CLI

Building a Roguelike from Your Codebase with GitHub Copilot CLI

A developer turned his codebase into a terminal roguelike using GitHub Copilot CLI's /delegate command and Binary Space Partitioning. Same commit, same dungeon. New commit, new layout. Here's how procedural generation and async AI coding actually work in practice.

Your First Open Source Contribution: A Practical Guide for Devs

Your First Open Source Contribution: A Practical Guide for Devs

Use GitHub Copilot Chat to find beginner-friendly open source projects, then fork, branch, and submit your first pull request. Here's the exact workflow GitHub recommends for new contributors.

ReSharper 2026.2 Brings Any AI Agent to Visual Studio via ACP

ReSharper 2026.2 Brings Any AI Agent to Visual Studio via ACP

ReSharper 2026.2 EAP introduces ACP support, letting you run any compatible AI agent in Visual Studio. Junie is first — autonomous coding, refactoring, and Git ops. The Agent Registry is coming.

GitHub Data Reveals Which Countries Lead in Software Complexity

GitHub Data Reveals Which Countries Lead in Software Complexity

Researchers used GitHub data to measure "software economic complexity" across 163 countries. Germany, Australia, and Canada lead. The findings predict GDP and inequality better than traditional trade data — and reveal which countries are positioned to win in the AI era.

How GitHub Cut Agentic Workflow Costs by 62% With Token Optimization

How GitHub Cut Agentic Workflow Costs by 62% With Token Optimization

GitHub reduced token usage in production agentic workflows by up to 62% through systematic optimization. Here's how they instrumented token consumption, built self-optimizing workflows, and cut costs without sacrificing quality.

Agent Pull Requests Are Everywhere. Here's How to Review Them.

Agent Pull Requests Are Everywhere. Here's How to Review Them.

Agent-generated code looks clean but introduces more technical debt than human code. Here's a 10-minute review protocol to catch CI gaming, code duplication, and hallucinated correctness before they ship.