Conduct periodic access reviews and certifications. Implement access governance and recertification workflows. Use when managing access compliance.
日本語の概要は準備中です。原文の説明を表示しています。
Audit and remediate CIS benchmark violations. Use automated tools to assess compliance and implement hardening recommendations. Use when meeting compliance requirements or implementing security baselines.
インストールする前に、エージェントに与えられる指示の中身を確認できます。
Implement and audit CIS security benchmarks.
Use this skill when:
# Install
apt install openscap-scanner scap-security-guide
# Run CIS benchmark scan
oscap xccdf eval \
--profile xccdf_org.ssgproject.content_profile_cis \
--results results.xml \
--report report.html \
/usr/share/xml/scap/ssg/content/ssg-ubuntu2204-ds.xml
# Install
apt install lynis
# Run audit
lynis audit system
# Generate report
lynis audit system --report-file /tmp/lynis-report.dat
# cis-profile/controls/ssh.rb
control 'cis-ssh-1' do
impact 1.0
title 'Ensure SSH root login is disabled'
describe sshd_config do
its('PermitRootLogin') { should eq 'no' }
end
end
control 'cis-ssh-2' do
impact 0.7
title 'Ensure SSH password authentication is disabled'
describe sshd_config do
its('PasswordAuthentication') { should eq 'no' }
end
end
# Run InSpec
inspec exec cis-profile -t ssh://user@target
# kube-bench
docker run --rm -v /etc:/etc:ro -v /var:/var:ro \
aquasec/kube-bench:latest run --targets node
# Check specific sections
kube-bench run --targets master --check 1.1,1.2
workflow:
1_scan:
- Run automated assessment
- Generate baseline report
2_analyze:
- Review findings
- Identify false positives
- Prioritize by risk
3_remediate:
- Apply fixes
- Document exceptions
- Verify changes
4_validate:
- Re-run assessment
- Confirm remediation
- Generate compliance report
まだレビューはありません。使ってみた感想をお寄せください。
概要と使いどころ
Conduct periodic access reviews and certifications. Implement access governance and recertification workflows. Use when managing access compliance.
日本語の概要は準備中です。原文の説明を表示しています。
Build automated evaluation suites for AI agents using golden datasets, rubrics, and regression gates. Use when shipping agent features, validating prompt changes, or gating deployments on quality.
日本語の概要は準備中です。原文の説明を表示しています。
Instrument AI agents with tracing, token metrics, latency, and cost visibility. Use for reliability and debugging.
日本語の概要は準備中です。原文の説明を表示しています。
Secure AI agents against prompt injection, tool abuse, and data exfiltration with defense-in-depth controls. Use when building, deploying, or hardening agentic AI systems that invoke tools, access data, or interact with production infrastructure.
日本語の概要は準備中です。原文の説明を表示しています。
Secure AI coding agents (Claude Code, Cursor, Codex, Copilot) with permission boundaries, secret protection, code review gates, and safe sandbox configurations for team environments.
日本語の概要は準備中です。原文の説明を表示しています。
Use service mesh patterns for AI inference traffic management, mTLS, canary releases, policy enforcement, and cross-cluster resilience.
日本語の概要は準備中です。原文の説明を表示しています。