google-cloud-scc-remediation

Remediates Google Cloud Security Command Center findings across IAM, infrastructure, and vulnerability categories.

23.7k|2.0k|Updated Mar 31, 2026
Ready to run
npx skills add https://github.com/google/skills --skill google-cloud-scc-remediation-google
💬 Prompt for ChatGPT / Claude Chat▼
Please help me install this Agent Skill.
Skill: google-cloud-scc-remediation
Source: https://github.com/google/skills/tree/main/skills/cloud/google-cloud-scc-remediation
Command: npx skills add https://github.com/google/skills --skill google-cloud-scc-remediation-google

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve? Security findings pile up faster than teams can fix them. This Skill automates the remediation of Security Command Center findings—from IAM permission fixes to firewall misconfigurations to vulnerable packages—while enforcing safety gates and verification steps so you never accidentally break production. ## Core Features & Use Cases - IAM Remediation: Fix over-privileged service accounts, revoke basic roles, rotate unmanaged keys, and enforce separation of duties. - Infrastructure Fixes: Restrict public bucket access, tighten firewall rules, remove external IPs, and configure KMS key rotation. - Vulnerability Patching: Upgrade vulnerable GKE node pools, apply OS patches to Compute Engine VMs, and rebuild container images with patched base images. - Use Case: A Security Command Center finding flags your production Cloud Storage bucket as publicly accessible via allUsers. Present the remediation plan (enable Public Access Prevention, remove the public IAM binding), get approval, execute, and verify the bucket is now private—all in a structured workflow. ## Quick Start Provide the Security Command Center finding details, and the skill will analyze the category, load the matching remediation playbook, present a structured plan with verification and rollback commands, and execute only after your explicit approval.

Frequently Asked Questions about google-cloud-scc-remediation

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

FAQPage Schema
How do I remediate an over-privileged service account in Security Command Center?▼

The skill identifies the overly broad basic role, grants the least-privilege replacement role first to prevent outage, then revokes the basic role. It presents the exact gcloud commands, verification steps, and rollback procedure before execution, requiring your explicit approval.

What happens if I have a publicly accessible Cloud Storage bucket finding?▼

The skill enables Public Access Prevention on the bucket and removes all public IAM bindings (allUsers and allAuthenticatedUsers). It verifies both steps succeeded before marking remediation complete.

Does the skill handle Toxic Combinations across multiple finding categories?▼

Yes. For multi-domain attack paths, the skill identifies each exposed facet and loads matching reference playbooks sequentially, presenting a comprehensive remediation plan that addresses all interconnected risks.

Why does the skill require explicit approval before executing remediation commands?▼

Mutating infrastructure, access policies, or workload deployments can cause outages or lockouts if the wrong resource is targeted. The skill presents the full plan, exact commands, verification checks, and rollback procedures so you confirm the impact before execution.

What if a Security Command Center finding doesn't have a dedicated playbook?▼

For runtime threat detections and findings without automation playbooks, the skill provides documentation, manual investigation steps, containment guidance, and links to relevant GCP documentation instead of attempting automated mutation.

How does the skill verify that a remediation actually worked?▼

After each mutating command, the skill runs verification commands (like gcloud describe or get-iam-policy) to confirm the resource is in the desired state. If verification fails, execution stops and the error is reported.