A peer-reviewed state-of-the-art review of vibe coding assembles 17 months of empirical evidence and finds the productivity numbers contradict each other until you control for scope and time horizon.

Laravel Auditor is a free MIT-licensed dev package that equips Claude, Cursor, Codex, and other agents with a structured methodology for auditing Laravel 12 and 13 apps.

The GoLand team released Modern Go Guidelines, a free CLI plugin that teaches AI agents to write version-matched Go code from Go 1.0 through Go 1.27.

Slack is rolling out Slack Code, dedicated project channels where Claude Code, Devin, Vercel Agent, and GitHub Copilot work alongside your team. Available on every Slack plan.

Salesforce launched Slack Code, a service that runs coding agents like Claude, Devin, and GitHub Copilot inside dedicated Slack channels so all stakeholders can watch, steer, and sign off in one place.

herdr is a terminal tool built to manage multiple AI coding agents at once, showing you which agents are waiting for input without tab-switching.

AI coding agents generate 12+ model calls and 80K+ tokens on tasks where a chat assistant uses one call and 5K tokens. Here is how to measure the real difference.

A production study of 761 million GitHub Copilot LLM calls found KV-cache reuse at 90% within a turn but only 55% across turn boundaries. Here is what that means for how you build agents.

Slack launched Slack Code, project-specific channels designed for teams to collaborate with AI coding agents across all plans. Here is what changed.

Ahmet Rıfat Öztürk shipped BrowserAgent, a Python 3.11+ MIT licensed toolkit for AI social media posting, browser automation, and a web-based Agent Studio panel.

Laravel Tackle is an MIT-licensed AI coding agent that runs as Artisan commands in Laravel 12 and 13, with PHP-enforced guardrails, self-healing queue jobs, and MCP support. Currently at v1.27.3.

One developer argues AI coding agents keep hitting architectural drift. Their proposed fix: an Architecture Graph, formal validation, and a deterministic generator between the two LLM calls.