crow-pet-policy-gate

Evaluate action proposals for policy compliance across scope, sensitivity, and blast radius.

Updated Mar 22, 2026
One-click install
npx skills add https://github.com/infolog-io/crow.pet --skill crow-pet-policy-gate
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: crow-pet-policy-gate
Source: https://github.com/infolog-io/crow.pet/tree/main/skills/crow-pet-policy-gate
Command: npx skills add https://github.com/infolog-io/crow.pet --skill crow-pet-policy-gate

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) and references (resource) components.

What problem does it solve?

This Skill helps classify and manage risky or sensitive actions in AI-powered systems, preventing unsafe operations and ensuring policy compliance.

Core Features & Use Cases

  • Decision Classification: Analyzes proposed actions on axes like scope and sensitivity to assign a decision level.
  • Policy Enforcement: Ensures that external writes, secrets, and destructive actions are properly vetted before execution.
  • Use Case: When an AI plans to push code to a repository or access sensitive data, this Skill evaluates the risk and recommends whether to proceed, ask for permission, notify, or block.

Quick Start

Request the policy classifier to evaluate a proposed action involving external system changes or sensitive data access.

Frequently Asked Questions about crow-pet-policy-gate

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

FAQPage Schema
How do I enforce security policy compliance before an AI workflow executes destructive actions?

To enforce security policy compliance, you can evaluate proposed AI actions against policy rubrics to classify their risk level and determine whether to block, notify, or proceed. This vetting prevents unauthorized external writes and destructive operations before execution.

What is policy-based decision classification for AI automation workflows?

Policy-based decision classification analyzes proposed AI actions across axes like scope, sensitivity, and blast radius to assign a risk level. It uses policy and rubric files to ensure consistent, safe decision-making for sensitive operations involving secrets or external changes.

How do I evaluate the blast radius of an AI action that accesses sensitive data?

You evaluate the blast radius by classifying the action proposal against security and sensitivity rubrics. This assessment assigns a decision level that recommends whether to ask for permission, block, or allow the operation based on its potential impact on sensitive data.

Can I use policy rubrics to gate external writes in automated AI systems?

Yes, you can use policy rubrics to gate external writes by evaluating the proposed action's scope and sensitivity. The evaluation generates a decision to proceed, ask for permission, notify, or block, ensuring external writes are properly vetted for compliance.

Do I need custom policy files to manage risk in AI decision-making workflows?

Yes, this approach relies on policy and rubric files to maintain consistency and safety during risk management. These files define the security, safety, and compliance standards used to evaluate action proposals and classify their decision levels.