cubic-audit

Audit cubic.yaml rules and reconcile learnings into actionable cleanup recommendations.

4.0k|550|Updated Feb 27, 2019
One-click install
npx skills add https://github.com/cartography-cncf/cartography --skill cubic-audit
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: cubic-audit
Source: https://github.com/cartography-cncf/cartography/tree/main/.agents/skills/cubic-audit
Command: npx skills add https://github.com/cartography-cncf/cartography --skill cubic-audit

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Audit and improve the cubic AI-review configuration for this repo by validating rule relevance, scope, and performance, and by reconciling them with live cubic learnings to maintain high-quality automation.

Core Features & Use Cases

  • Reviews the cubic.yaml configuration for relevance, scope, limits, and factual accuracy against the codebase.
  • Reconciles accumulated cubic learnings into durable rules while flagging obsolete, duplicate, or contradictory learnings for cleanup.
  • Produces an actionable plan with deliverables for rule reviews and learnings reconciliation across both repo and organizational contexts.

Quick Start

Run cubic-audit to perform a full audit of cubic.yaml and associated learnings, then review the recommended promotions and cleanups.

Frequently Asked Questions about cubic-audit

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

FAQPage Schema
How do I audit and clean up cubic AI-review rules for my repository?

To audit cubic AI-review rules, validate the cubic.yaml structure, check rule relevance against your codebase, and reconcile accumulated learnings into durable, high-signal automation configurations.

What is the best way to reconcile cubic learnings into durable rules?

Reconciling cubic learnings involves reviewing accumulated data to promote useful patterns into rules while flagging obsolete, duplicate, or contradictory learnings for cleanup to maintain high-quality automation.

Can I validate cubic.yaml rule scope and relevance without the cubic MCP connected?

You can validate cubic.yaml without the cubic MCP connected; the audit performs a rules-only pass, checking YAML structure and factual accuracy against the codebase instead of a full audit.

Does the cubic audit work with organizational contexts or only single repositories?

The cubic audit works across both repository and organizational contexts, producing an actionable plan with deliverables for rule reviews and learnings reconciliation tailored to each scope.

Why are my cubic AI-review rules generating low-signal or contradictory results?

Low-signal or contradictory cubic AI-review results often stem from obsolete or duplicate learnings; auditing the cubic.yaml configuration identifies these issues and outputs cleanup recommendations.