Security operations with Claude Code — 32 specialized plugins plus an optional hooks plugin, covering DevSecOps, threat modeling, incident response, authorized penetration testing, and cloud security
Skills, agents, commands, and workflows for security operations with Claude Code.
Security work is asymmetric. Defenders must be right every time. Attackers need to be right once. Generic AI coding produces "looks secure" code that fails real adversarial review. LibreSecOps gives Claude Code the security-domain expertise needed to ship systems that survive real attacks.
Thirty-two domain plugins covering blue team, red team, cloud security, application security, compliance, and the operational layer that sits between them.
Karpathy, December 2025: programming is being refactored. For security specifically, AI-generated code introduces new vulnerability surfaces — prompt injection, supply-chain risks, misuse of cryptographic primitives, over-permissive IAM. The defenders' toolkit has to evolve with the threats.
| Claude Code component | LibreSecOps provides |
|---|---|
| Plugins | 32 subdomain plugins (threat modeling, IR, pentesting, cloud sec, app sec, compliance, more) plus the optional libre-secops-hooks plugin |
| Agents | 64 specialist agents, two per plugin (threat modeler, incident commander, pentest planner, detection engineer, and more) |
| Commands | 53 slash commands, namespaced by plugin (for example /threat-modeling:threat-model) |
| Skills | 57 pattern libraries (STRIDE, MITRE ATT&CK mappings, OWASP categories, NIST controls) |
| Hooks | Optional: a security profile of the project at session start, a confirmation prompt before edits to secret files, and a vulnerability pattern scan after each edit |
| Templates | Threat model templates, IR playbooks, audit-evidence scaffolds |
| Plugin | Domain | Agents |
|---|---|---|
| threat-modeling ⭐ | STRIDE, attack trees, MITRE ATT&CK mapping | attack-tree-builder, threat-modeler |
| blue-team-detection | Detection engineering, alerting, SIEM rule design | detection-engineer, threat-hunter |
| incident-response | IR playbooks, containment, forensics handoff | forensic-analyst, incident-commander |
| siem-log-management | Log normalization, alert tuning, threat hunting | log-analyst, siem-architect |
| forensics-analysis | Digital forensics, evidence chain, memory analysis | digital-forensics-examiner, memory-forensics-analyst |
| malware-analysis | Static + dynamic analysis, sandboxing, IoC extraction | dynamic-analyst, static-analyst |
| security-automation | SOAR, playbook automation, response orchestration | automation-builder, soar-architect |
| security-awareness | Phishing training, user education, social engineering defense | policy-writer, security-trainer |
| social-engineering-defense | Anti-phishing, anti-pretexting, anti-vishing | awareness-program-designer, social-engineering-analyst |
| Plugin | Domain | Agents |
|---|---|---|
| penetration-testing | Pentest methodology (PTES, OWASP Testing Guide, NIST SP 800-115), scoping, rules of engagement, reporting | pentest-planner, vuln-researcher |
| red-team-operations | Adversary emulation planning, rules of engagement, MITRE ATT&CK mapping, detection validation | red-team-lead, ttp-researcher |
| bug-bounty-methodology | Recon, vulnerability discovery, responsible disclosure | bug-bounty-hunter, vuln-report-writer |
| vulnerability-scanning | Scanner selection (Nessus, Qualys, OpenVAS), false-positive triage | vuln-scanner-orchestrator, vuln-triager |
| api-security-testing | OWASP API Top 10: BOLA, BFLA, mass assignment, auth flows, GraphQL-specific | api-security-tester, auth-flow-auditor |
| web-application-security | OWASP Top 10, XSS, SQLi, CSRF, authentication flaws | web-security-auditor, xss-hunter |
| Plugin | Domain | Agents |
|---|---|---|
| cloud-security-aws | AWS IAM, KMS, GuardDuty, Security Hub, well-architected security pillar | aws-compliance-auditor, aws-security-architect |
| cloud-security-azure | Azure AD, Defender, Sentinel, Conditional Access | azure-compliance-auditor, azure-security-architect |
| cloud-security-gcp | GCP IAM, Security Command Center, Organization Policies, VPC Service Controls | gcp-org-policy-auditor, gcp-security-architect |
| container-security | Image scanning, runtime security, SBOM, Distroless | container-hardener, image-scanner |
| kubernetes-security | Pod Security Standards, RBAC, NetworkPolicies, OPA Gatekeeper | k8s-policy-enforcer, k8s-security-architect |
| Plugin | Domain | Agents |
|---|---|---|
| secure-coding-practices | Language-specific anti-patterns (SQL injection, deserialization, etc.) | input-validation-specialist, secure-code-reviewer |
| supply-chain-security | SBOM, SLSA, dependency confusion, typosquatting defense | dependency-auditor, package-integrity-analyst |
| devsecops-pipelines | Shift-left security, SAST/DAST/SCA in CI | devsecops-architect, pipeline-security-integrator |
| cryptography-essentials | Symmetric vs asymmetric, key management, common mistakes (ECB, IV reuse) | crypto-advisor, tls-specialist |
| secrets-management | Vault, AWS Secrets Manager, GCP Secret Manager, rotation patterns | secret-scanner, secrets-architect |
| mobile-app-security | OWASP MASVS and MASTG, certificate pinning, keychain security | mobile-code-reviewer, mobile-security-tester |
| Plugin | Domain | Agents |
|---|---|---|
| identity-access-management | RBAC, ABAC, OAuth2, OIDC, SAML, just-in-time access | access-control-auditor, iam-architect |
| zero-trust-architecture | Beyond perimeter, identity-first networking, microsegmentation | microsegmentation-specialist, zero-trust-architect |
| privacy-engineering | GDPR, CCPA, data minimization, privacy-by-design patterns | dpia-analyst, privacy-engineer |
| Plugin | Domain | Agents |
|---|---|---|
| compliance-frameworks | SOC 2, ISO 27001, PCI DSS, HIPAA, NIST CSF mappings | compliance-auditor, evidence-collector |
| security-hardening | CIS benchmarks, host hardening, network hardening, OS-specific configs | benchmark-auditor, hardening-specialist |
| network-security | Firewall design, segmentation, IDS/IPS, DNS security | ids-ips-engineer, network-security-architect |
⭐ = flagship plugin, the first taken to full depth (see the 0.2.0 entry in the CHANGELOG). Every plugin ships two agents, one or two slash commands, and one to three skills.
| Plugin | What it does |
|---|---|
| libre-secops-hooks | At session start, one line of security context about the project (stack, auth libraries, infrastructure, security tooling, gaps such as a .env tracked by git). Before an edit to a .env, key, or credentials file, a confirmation prompt. After each edit, a local pattern scan for injection, XSS, hardcoded secrets, weak crypto, and Dockerfile issues. See its README. |
/plugin marketplace add HermeticOrmus/LibreSecOps-Claude-Code
/plugin install threat-modeling@libre-secops
Install any other plugin the same way (/plugin install <plugin>@libre-secops), or open /plugin to browse the pack. From a terminal, the equivalent is:
claude plugin marketplace add HermeticOrmus/LibreSecOps-Claude-Code
claude plugin install threat-modeling@libre-secopsThe optional hooks plugin installs the same way: /plugin install libre-secops-hooks@libre-secops (it needs jq; see its README). Restart Claude Code after installing.
Grok Build reads the same plugin folders. Add the marketplace, then install any plugin by name:
grok plugin marketplace add HermeticOrmus/LibreSecOps-Claude-Code
grok plugin install threat-modeling@LibreSecOps-Claude-Code --trustOr install one plugin straight from its folder, without adding the marketplace:
grok plugin install HermeticOrmus/LibreSecOps-Claude-Code#plugins/threat-modeling --trust--trust confirms you trust the source; without it Grok shows what the plugin would activate and stops. Start a new Grok session to load what you installed. From a clone, ./setup.sh --grok installs the whole pack through the grok CLI. The libre-secops-hooks plugin uses a hook format Grok supports, but it has not been verified in a live Grok session (see the ledger).
git clone https://github.com/HermeticOrmus/LibreSecOps-Claude-Code.git ~/projects/LibreSecOps-Claude-Code
cd ~/projects/LibreSecOps-Claude-Code
./setup.shsetup.sh registers the checkout as a plugin marketplace and installs every plugin through the Claude Code CLI, including libre-secops-hooks. It needs claude and jq on your PATH. Useful flags: --list shows the plugins, --only threat-modeling,incident-response installs a subset (leave libre-secops-hooks out of the list to skip the hooks), --scope project installs for the current project only, and --uninstall removes the pack. Add --grok to install through Grok Build instead; it works with --list, --only, and --uninstall, and needs grok and jq.
Then in any Claude Code session:
/threat-modeling:threat-model build a STRIDE threat model for a SaaS application with multi-tenant data, OAuth2 social login, file upload, and a public REST API
Plugin commands are namespaced as /<plugin>:<command>, so the /threat-model command in the plugin docs runs as /threat-modeling:threat-model. Agents are chosen automatically from their descriptions, or you can call one directly, for example @agent-incident-response:incident-commander.
See QUICK_START.md for the full walkthrough.
- Beginner — security mindset shifts, your first threat model, OWASP Top 10
- Intermediate — DevSecOps integration, IR playbooks, cloud security posture
- Advanced — red team / blue team exercises, compliance audit prep, zero-trust migration
- Cloud platforms: AWS, Azure, GCP (parity across the three)
- Container/orchestration: Docker, Kubernetes, ECS, GKE, AKS
- Compliance frameworks: SOC 2, ISO 27001, PCI DSS, HIPAA, NIST CSF, FedRAMP, GDPR
- Languages: Python, TypeScript, Go, Rust, Java, .NET
- Skill level: developers entering security through senior security engineers
- Claude Code: installs as a plugin marketplace (tested with Claude Code 2.1.285); the hooks plugin needs
bashandjq - Grok Build: installs as a plugin marketplace (tested with grok 1.0.44); the hooks plugin has not been verified in a live Grok session
This kit is for defensive security and authorized testing only. The offensive plugins (penetration-testing, red-team-operations, bug-bounty-methodology) are intended for use with explicit authorization. Unauthorized testing is illegal in most jurisdictions.
This is documentation + prompt-engineering. It is not:
- Legal advice on compliance
- A replacement for certified security professionals
- An audit certification
For regulated systems, retain licensed security counsel and accredited auditors.
Starred this? Tell us what worked and what is missing: open a feedback issue. Every piece of feedback gets an answer, and changes that come from it are credited in the release notes.
Cracks we found and sealed: LEDGER.md.
- Take the next piece of work from the Menu, or start with a good first issue.
- Claude picked the wrong agent or skill? File a routing miss.
- Have a plugin in mind? Open a plugin proposal. Anything else goes in the feedback form.
- Share threat models, detection rules and playbooks in Discussions.
How to claim an item and test a change locally: Ways to contribute. Report a vulnerability in this repo through Security Advisories, not a public issue.
PRs welcome — especially: more depth on cloud security per platform, regional compliance translations (LATAM LGPD, India DPDP, etc.), real-world incident case studies (anonymized), supply-chain attack postmortems.
See CONTRIBUTING.md.
MIT.
This repository is part of a growing family of open-source toolkits for Claude Code.
- LibreUIUX-Claude-Code — UI/UX development (152 agents, 70 plugins, 76 commands, 74 skills)
- LibreArch-Claude-Code — Software architecture and system design
- LibreCopy-Claude-Code — Technical writing and documentation engineering
- LibreDevOps-Claude-Code — DevOps engineering and infrastructure automation
- LibreEmbed-Claude-Code — Embedded systems, firmware, and IoT development
- LibreFinTech-Claude-Code — Financial technology development
- LibreGEO-Claude-Code — AI-search optimization (ChatGPT, Perplexity, Gemini, Google AI Overviews)
- LibreGameDev-Claude-Code — Game development across Godot, Unity, Unreal
- LibreMLOps-Claude-Code — ML engineering and AI operations
- LibreMobileDev-Claude-Code — Mobile app development (Flutter, React Native, native iOS, native Android)
- LibreSessionFlow-Claude-Code — Session lifecycle: handoff, pickup, absorb, explore, close
- vibe-engineer-skills — Direct AI codegen well: hypothesis before help, scoped prompts, validate before accepting
- markdown-discipline-skills — Strip AI-slop from markdown (no em dashes, no marketing fluff)
- shell-safety-skills —
set -euo pipefaildiscipline plus 15 failure-mode examples - commit-standard-skills — Ormus Commit Standard v1.0 plus commit-msg hook and commitlint
- unwoke-skills — Strip AI theater (ten sins to eliminate, symmetric engagement)
- python-conventions-skills — Modern Python 3.11+ (types, pathlib, async, ruff, mypy, uv)
- typescript-conventions-skills — TypeScript strict mode, discriminated unions, Result types
- hermetic-laws-skills — Seven Hermetic Principles applied to engineering
- riper-workflow-skills — Research / Innovate / Plan / Execute / Review systematic dev
- six-day-cycle-skills — Sustainable shipping cadence with mandatory rest
- token-optimization-skills — Claude Code token and context optimization
- osint-skills — OSINT research methodology (multi-wave investigative spiral)
- calcinate-skills — Stage 1 of the Magnum Opus (burn project bloat)
- claude-md-overhaul-skills — Audit CLAUDE.md and MEMORY.md against caps
- session-handoff-skills — Session handoff and pickup discipline
- naming-skills — Product naming methodology (mine the brand's vocabulary)
- magnum-opus-skills — Seven-stage alchemy applied to project transformation
- mem-search-skills — Search claude-mem cross-session memory: search, filter, fetch
- hypothesis-debugging-skills — Hypothesis-driven debugging: reproduce, isolate, test, fix
- vibe-proof-skills — Security hardening for vibe-coded full-stack apps
- tdd-skills — Test-driven development (Red-Green-Refactor) for JS/TS and Python
- mars-skills — Production-readiness audit: the five mortal sins of vibe-coded MVPs
- git-workflow-skills — Clean git workflow: branch, atomic commits, reviewable PRs
- code-review-skills — Domain-aware code review: classify the code, then focus
- code-comprehension-skills — Understand an unfamiliar codebase fast
- dx-audit-skills — Audit developer experience: docs, onboarding, tooling friction
- setup-env-skills — Set up a project's development environment
- automate-skills — Turn repetitive tasks into reliable automation scripts
- quick-fix-skills — Fast troubleshooting for common issues
- prime-context-skills — Prime project context at the start of a session
- auto-docs-skills — Generate and maintain project documentation
- learning-skills — Learn any technology: roadmaps, explanations, practice, cheatsheets, comparisons
- linux-sysadmin-skills — Linux system administration: security, performance, diagnostics, monitoring, maintenance
- andrej-karpathy-skills — the canonical single-file CLAUDE.md pattern (fork of jiayuan_jy's original)
Star the family, not just one — that's how the suite stays coherent.