guard

Enforce directory-scoped edits and destructive-command warnings during sensitive operations.

1|Updated Feb 25, 2026

Search Tags:#devops #safety #freeze #guard #destructive-warnings #edit-boundary #careful

Ready to run

npx skills add https://github.com/alexandervashurin/semaphore --skill guard-alexandervashurin

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: guard
Source: https://github.com/alexandervashurin/semaphore/tree/main/.claude/skills/guard
Command: npx skills add https://github.com/alexandervashurin/semaphore --skill guard-alexandervashurin

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Full safety mode activates destructive-command warnings and directory-scoped edits to protect production and live debugging environments, combining the capabilities of the /careful and /freeze skills.

Core Features & Use Cases

Destructive-command warnings for risky actions (rm -rf, DROP TABLE, force-push, etc.) before execution.
Edit-boundary enforcement to restrict edits to a user-specified directory.
Interactive setup prompts to select the target directory and establish the freeze boundary.
Dependency-aware integration with sibling /careful and /freeze skills for comprehensive safety.
Safe defaults and clear user messaging to prevent accidental damage during critical operations.

Quick Start

Use guard mode to enable dual protections and restrict edits to a specific directory while warning on destructive commands.

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 production debugging?▼

Preventing destructive commands during production debugging requires safety mode warnings for actions like rm -rf or DROP TABLE before execution. This mechanism intercepts risky inputs and enforces directory-scoped edit boundaries to restrict changes.

What is directory-scoped edit restriction and how does it protect live environments?▼

Directory-scoped edit restriction protects live environments by enforcing a boundary that limits modifications to a user-specified directory. This prevents accidental file changes outside the defined scope during sensitive high-risk deployments or live debugging sessions.

How do I set up edit boundaries and destructive-command warnings for high-risk deployments?▼

Setting up edit boundaries and destructive-command warnings for high-risk deployments involves using interactive setup prompts to select a target directory. This establishes a freeze boundary and activates dual protections against risky actions.

Do I need sibling /careful and /freeze skills to use full safety mode?▼

Yes, full safety mode depends on sibling /careful and /freeze skills for setup and validation. It combines their destructive-command warnings and directory-scoped edit restrictions to provide comprehensive dependency-aware integration during critical operations.

What are the limitations of relying on edit-boundary enforcement for prod debugging?▼

The limitation of edit-boundary enforcement for prod debugging is that it restricts edits strictly to a user-specified directory. Operations requiring cross-directory modifications or force-pushes outside the freeze boundary will be blocked by the safety constraints.