safety-guardrails

Enforce safety checks for destructive operations and edit-scope expansion.

1|Updated Jul 9, 2026
One-click install
npx skills add https://github.com/PiercingXX/xx-stack --skill safety-guardrails
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: safety-guardrails
Source: https://github.com/PiercingXX/xx-stack/tree/main/runtime/skills/safety-guardrails
Command: npx skills add https://github.com/PiercingXX/xx-stack --skill safety-guardrails

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This skill reduces operational risk during sensitive work by forcing clarity and explicit confirmation before destructive or wide-scope changes.

Core Features & Use Cases

  • Destructive Command Guardrails: Requires confirmation and impact disclosure for risky operations like recursive deletes, hard resets, force pushes, and production deletions.
  • Edit Boundary Discipline: Keeps edits scoped to the user-provided target folder and asks first when changes must occur outside scope.
  • Failure Classification & Circuit Breakers: Classifies risk by determinism, environment readiness, transience, or operator capability, and stops repeating blocked risky attempts until rollback/scope evidence is provided.

Quick Start

Use the safety-guardrails skill whenever the requested change can delete data, mutate history, or expand edit scope, and ask it to output a risk level, approval decision, rollback plan, and verification state before proceeding.

Frequently Asked Questions about safety-guardrails

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

FAQPage Schema
How do I prevent destructive operations like force pushes and recursive deletes in AI-assisted development?

Preventing destructive operations requires enforcing manual approval gates and impact disclosure before executing risky commands like force pushes, recursive deletes, or hard resets. This ensures explicit confirmation happens prior to any irreversible impact.

What is edit boundary discipline and how does it control risky edit-scope expansion?

Edit boundary discipline restricts modifications to a user-provided target folder and requires explicit confirmation before expanding the edit scope. This mechanism prevents unauthorized changes to files outside the defined operational boundary.

How do I add a circuit breaker for repeated blocked risky attempts during rollback planning?

To add a circuit breaker, classify operation risks by environment readiness and operator capability, then halt repeated blocked risky attempts until rollback and scope evidence is provided. This stops infinite loops of failed destructive actions.

Does this safety guardrails approach work for download-and-execute workflows and history mutations?

Yes, safety guardrails apply to download-and-execute workflows and history mutations by classifying determinism and transience risks. These operations require manual approval gates and rollback evidence checks before proceeding.

What are the limitations of using safety guardrails for production deletions and environment classification?

Safety guardrails require manual intervention for production deletions, meaning automated pipelines will stall until a human provides approval. They also depend on accurate environment and capability classification to function correctly.