guard

Aggregate destructive command warnings and edit-boundary protections for prod systems.

Updated Mar 15, 2026
One-click install
npx skills add https://github.com/lachieh/babytalk --skill guard-lachieh
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: guard
Source: https://github.com/lachieh/babytalk/tree/main/.claude/skills/gstack/guard
Command: npx skills add https://github.com/lachieh/babytalk --skill guard-lachieh

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Full safety mode addressing risk of destructive commands and unintended edits by enforcing a combined guard that warns on destructive operations and blocks edits outside a designated directory.

Core Features & Use Cases

  • Combines /careful destructive-command warnings with /freeze edit-boundary enforcement to protect prod systems.
  • Asks the user to specify a directory to restrict edits to, then activates two-layer protection: warnings for dangerous commands and a hard edit boundary.
  • Integrates with existing gstack workflows and can be installed alongside careful and freeze.

Quick Start

Ask the user which directory should be protected and then enable guard mode with destructive-command warnings active.

Frequently Asked Questions about guard

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I prevent accidental destructive commands during prod maintenance?

To prevent destructive commands during prod maintenance, you can enable a safety mode that aggregates warnings for risky operations and blocks unintended edits outside a designated directory.

Can I restrict edits to a specific directory while debugging production systems?

Yes, you can restrict edits to a specific directory while debugging production systems by activating a freeze boundary that enforces a hard edit boundary, preventing modifications outside the protected path.

How does combining destructive command warnings with edit boundaries work?

Combining destructive command warnings with edit boundaries works by applying a two-layer protection that alerts on dangerous operations and enforces a hard block on edits outside a specified directory.

What is the best way to protect prod systems from unintended edits and risky commands?

The best way to protect prod systems is activating a full safety mode that integrates careful destructive-command warnings with freeze edit-boundary enforcement, storing the configuration centrally for the active session.

Do I need to specify a directory to enable full safety mode for risky commands?

Yes, you need to specify a directory to restrict edits to, which then activates two-layer protection combining destructive-command warnings and a hard edit boundary for the active session.

When should I not use a combined guard for destructive operations?

You should not use a combined guard when performing operations requiring edits outside the designated directory, as the hard edit boundary will actively block modifications beyond the specified path.