Top AI Product

Every coding agent user shares the same nightmare: the agent runs git reset --hard or rm -rf ./src and hours of uncommitted work vanish. destructive_command_guard (dcg), a safety hook by Jeff Emanuel, intercepts those commands before they execute. It added 444 stars in a single day and now sits at 2.5k, topping GitHub trending.

What it actually does

Written in Rust with SIMD-accelerated matching, dcg inspects every shell command an agent tries to run — in under a millisecond, most in under 100 microseconds. 50+ modular packs cover databases, Kubernetes, Docker, AWS/GCP/Azure, Terraform and CI. It even parses heredocs and inline scripts, so python -c "os.remove(...)" gets caught too. Works with Claude Code, Codex CLI, Gemini CLI, Cursor, Copilot CLI — basically every major agent, via one bash installer.

The smartest design call

It fails open. Timeout or parse error? The command passes. dcg would rather miss one bad command than break your workflow — the opposite instinct of most security tools, and exactly why developers actually keep it installed.

You Might Also Like

• Zcode Zhipu z ai glm 5 2 Coding Agent hit hns 1 Spot Chinas Open Answer to Claude Code
• 27k Github Stars in Weeks Learn Claude Code by Shareai lab Breaks Down ai Coding Agents Into 12 Lessons
• Claw Code Rewrote Claude Code in Rust Before Sunrise and hit 50k Github Stars in 2 Hours
• Hiveterm Bets on the Multi Agent Workspace Claude Codex and Gemini in one Terminal
• Deepseek tui Tops Github Trending a Claude Code Clone Wired to Deepseeks api