infra-audit

Audits live server security over SSH across twelve infrastructure dimensions with deterministic collection and parallel analysis.

6|Updated Mar 27, 2026
One-click install
npx skills add https://github.com/greglas75/zuvo --skill infra-audit-greglas75
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: infra-audit
Source: https://github.com/greglas75/zuvo/tree/main/skills/infra-audit
Command: npx skills add https://github.com/greglas75/zuvo --skill infra-audit-greglas75

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) components.

What problem does it solve? Manually checking the security posture of live servers is slow, inconsistent, and easy to get wrong. This Skill automates a read-only security audit of servers you own over SSH, covering SSH hardening, accounts, network exposure, TLS, firewall, patch posture, logging, web services, Docker, databases, filesystem, and host secrets, with an explicit authorization gate before any connection is made. ## Core Features & Use Cases - Deterministic collection plus LLM analysis: A collector script gathers normalized JSON bundles per host, then parallel analyst agents (host, network, container, data) interpret findings grounded in registry-defined check IDs and severities. - Dual-vantage network verification: Compares internal listeners against an external scan via proxy, scan-via SSH host, or direct mode to prove firewall effectiveness. - Safety gates: Per-target authorization confirmation, consent-gated tool installs, secret redaction at the collector, read-only operation, dry-run and resume support for interrupted fleet audits. - Use Case: Point it at a hosts.yaml inventory of your production fleet to receive per-host reports and a fleet summary with severity-ranked findings and CIS-mapped remediations. ## Quick Start Run the infra-audit skill with --host [email protected] to audit a single server you own, confirming the authorization gate when prompted.

Frequently Asked Questions about infra-audit

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

FAQPage Schema
How do I audit the security of my Linux server over SSH?

Run the skill with --host user@address for a single server or provide a hosts.yaml inventory for a fleet. It connects read-only over SSH, collects configuration evidence, and produces per-host reports with severity-ranked findings across twelve dimensions.

What security checks does a server infrastructure audit cover?

The audit covers twelve dimensions: SSH hardening, accounts and auth, network exposure, TLS certificates, firewall and kernel networking, patch posture, logging and intrusion detection, web services, Docker, database servers, filesystem hardening, and host secrets hygiene.

Is it safe to run a security audit against live production servers?

The audit is read-only with one consented exception for tool installation, which requires per-host approval. No SSH connection opens before an explicit authorization gate is confirmed, secrets are redacted at collection, and --dry-run prints all commands without connecting.

Can I resume an interrupted infrastructure audit run?

Yes, use --resume with the run directory to continue from state.json. The skill also auto-detects recent interrupted runs and skips hosts already reported, re-collecting only hosts with incomplete bundles.

How do I scan a server externally without installing tools on my laptop?

Use --scan-via with an SSH target to run the external leg from that remote host using portable nc, openssl, and curl. This is macOS-safe and avoids needing local nmap, testssl, or proxychains.