skills.rest
mindicator
mindicatorCOMMUNITY
Explore all skills by this author →
GitHub

invairiant

Audit codebases for architectural invariant violations across PRs, commit ranges, and modules.

29|1|Updated Jul 2, 2026
Search Tags:#AI coding agent#LLM#code audit#architecture audit#evidence-based
One-click install
npx skills add https://github.com/mindicator/invAIriant --skill invairiant
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: invairiant
Source: https://github.com/mindicator/invAIriant/tree/main/skill
Command: npx skills add https://github.com/mindicator/invAIriant --skill invairiant

SYSTEM DOCUMENTATION & REQUIREMENTS

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

What problem does it solve?

This Skill performs evidence-based architecture audits for AI-era codebases, ensuring architectural invariants remain stable under AI-assisted change.

Core Features & Use Cases

  • PR Audits: Conduct PR scoped audits with customizable lenses.
  • Full-scale Audits: Run comprehensive audits across all mandatory lenses for a full report.
  • Commit Range Audits: Analyze specific commit ranges for architectural drift.
  • Module Audits: Assess individual modules or directory subtrees.
  • ADR Audits: Review ADRs for code drift and compliance.
  • Refactoring Proposal Audits: Evaluate proposed refactorings for potential invariant violations.
  • Evidence Verification: Adversarially verify candidate findings.
  • Severity Classification: Classify findings into severity levels.
  • Report Synthesis: Assemble final audit reports.
  • Closure Verification: Re-verify fixed findings post-incident.

Quick Start

Run an audit on your current PR with the /invairiant audit-pr command.

Frequently Asked Questions about invairiant

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

FAQPage Schema
How do I perform an architecture audit on a pull request?▼

Perform an architecture audit on a pull request by running the `/invairiant audit-pr` command, which scopes the architectural review to your specific PR changes using customizable lenses.

What is an evidence-based code audit for AI-era codebases?▼

An evidence-based code audit adversarially verifies candidate findings to ensure architectural invariants remain stable under AI-assisted changes, classifying verified issues into specific severity levels.

Do I need a specific framework to run a commit range or module audit?▼

Yes, running commit range or module audits requires the invairiant framework and a properly configured environment to assess specific commits and directory subtrees for architectural drift.

Can I audit ADRs and refactoring proposals for invariant violations?▼

Yes, you can review ADRs for code drift and compliance, as well as evaluate proposed refactorings to detect potential architectural invariant violations before they are merged.

How do I verify that fixed architectural findings are actually resolved?▼

Verify resolved findings by running a closure verification step, which re-verifies fixed findings post-incident to ensure the architectural invariants are genuinely restored.

Related Skills

Explore More Software Engineering Skills→
HechaoYannetHechaoYannet

create-architecture

Generate a master architecture document from GDDs, ADRs, and governance constraints.

Community
Advanced
fluxforgeaifluxforgeai
3

finding

Log proactive discovery findings into classified Markdown reports and trackers.

Community
Intermediate
MohammedSaudAlsahliMohammedSaudAlsahli

karpathy

Apply four coding principles for cautious, minimal LLM-based code changes.

Community
Intermediate
ArabelaTsoArabelaTso
142

component-boundary-identifier

Analyze Python codebases to identify component boundaries and detect architectural violations.

Community
Intermediate