Blog
Thoughts on AI agent safety, production guardrails, and building production-safe agentic systems.
AI Agent Incidents This Week — Issue #1
A weekly roundup of real AI agent incidents from the community: what went wrong, root cause analysis, and the guardrails that would have prevented each one.
What Claude Code's Sandbox Actually Does (And Doesn't Do)
A technical deep dive into Claude Code's built-in safety mechanisms, their limitations, and how runtime guardrails fill the gaps.
The Complete Guide to Claude Code's --dangerously-skip-permissions Flag
Everything you need to know about --dangerously-skip-permissions: what it does, when to use it, the risks, and how to use it safely with Railguard.
Running 10 Claude Code Agents Simultaneously — Without Breaking Everything
How to run multiple Claude Code sessions in parallel safely. File-level locking, session coordination, and production guardrails with Railguard.
How to Use --dangerously-skip-permissions Safely in Claude Code
Claude Code's --dangerously-skip-permissions flag gives your agent full autonomy — but one wrong command can destroy production. Here's how Railguard makes it safe.
The Claude Code Terraform Destroy Incident — And How to Prevent It
On February 26, 2026, Claude Code ran terraform destroy on production and wiped 2.5 years of data. Here's what happened, why it keeps happening, and how Railguard stops it.
Why We Built Railguard
AI agents are fast. Railguard makes them production-safe — so you can run them at full speed without the 3am incident.
Claude Code in Production: How to Prevent the 3am Incident
Real incidents show what happens when AI coding agents run without guardrails. Here's how to make Claude Code production-safe with Railguard.
AI Agent Guardrails: Sandboxing vs Runtime Safety vs Manual Approval
Comparing approaches to AI agent safety — sandboxes, manual approval, and runtime guardrails like Railguard. Which one is right for Claude Code and AI coding agents?