{
  "markdown": "# Security Skills for AI Coding Agents\n\n**Drop structured security skills into your AI coding agent. Get instant, framework-grounded security expertise.**\n\n![License: MIT](https://img.shields.io/badge/License-MIT-blue.svg)\n![Skills: 45](https://img.shields.io/badge/Skills-45-green.svg)\n![Claude Code](https://img.shields.io/badge/Claude_Code-compatible-purple.svg)\n![Gemini CLI](https://img.shields.io/badge/Gemini_CLI-compatible-purple.svg)\n![Cursor](https://img.shields.io/badge/Cursor-compatible-purple.svg)\n![Codex CLI](https://img.shields.io/badge/Codex_CLI-compatible-purple.svg)\n![OpenClaw](https://img.shields.io/badge/OpenClaw-compatible-purple.svg)\n![Kiro](https://img.shields.io/badge/Kiro-compatible-purple.svg)\n\n---\n\n## Why This Exists\n\nAI coding agents can perform security reviews, but they hallucinate framework control numbers, miss entire vulnerability categories, and produce inconsistent output across runs. The result is security guidance that sounds authoritative but falls apart under scrutiny.\n\nThese skills ground agents in real published frameworks -- OWASP, NIST, MITRE ATT&CK, and CIS Controls -- so that every finding maps to a verifiable control. They are not prompt dumps. They are structured, framework-referenced, injection-hardened skill files that produce reliable, auditable security output.\n\n## Quick Start\n\n```bash\ngit clone https://github.com/UnitOneAI/SecuritySkills.git\ncd SecuritySkills\n```\n\n**Claude Code** (native format — auto-discovery and `/slash-commands`)\n\n```bash\n# Global install — all skills available via auto-discovery and /skill-name\ncp -r skills/*/* ~/.claude/skills/\n\n# Or project-local\nmkdir -p .claude/skills && cp -r skills/*/* .claude/skills/\n\n# Then use naturally:\n# \"Review this code for security issues\"    → Claude auto-loads secure-code-review\n# /threat-modeling                          → Direct invocation\n# /cve-triage CVE-2024-1234                 → With arguments\n```\n\n**Gemini CLI**\n\n```bash\n# Reference skills via @ commands\ncp -r skills/ ~/.gemini/skills/\n```\n\n**Cursor**\n\n```bash\n# Add as Cursor rules\ncp -r skills/ .cursor/rules/\n```\n\n**Codex CLI / Kiro / Generic**\n\n```bash\n# Point any agent at a skill's SKILL.md file\ncodex --context skills/appsec/threat-modeling/SKILL.md \"Review this design\"\nkiro spec --skill skills/ai-security/llm-top-10/SKILL.md\n```\n\nEach skill is a directory with `SKILL.md` as the entrypoint, following the [Agent Skills](https://agentskills.io) open standard. Claude Code discovers skills automatically; other tools can load them by path.\n\n## Skill format\n\nEvery skill is a **directory** at `skills/<domain>/<skill-name>/` with `SKILL.md` as the entrypoint, following the [Agent Skills](https://agentskills.io) open standard.\n\n### `SKILL.md` frontmatter\n\nAll skills use the same YAML frontmatter fields:\n\n```yaml\nname: threat-modeling                 # kebab-case, matches the directory\ndescription: >                        # what it does + when it auto-invokes\n  Runs a structured STRIDE threat model on any design, API spec, or codebase...\ntags: [appsec, design, architecture]  # domain + activity keywords\nrole: [security-engineer, architect]  # which role bundles include it\nphase: [design, review]               # SDLC phase\nframeworks: [STRIDE, MITRE-ATT&CK]    # cited frameworks — real control IDs only\ndifficulty: intermediate              # beginner | intermediate | advanced\ntime_estimate: \"30-60min\"\nversion: \"1.0.0\"\nauthor: unitoneai\nlicense: MIT\nallowed-tools: Read, Grep, Glob       # tools the skill may use\ninjection-hardened: true              # reviewed against OWASP LLM01\nargument-hint: \"[target-file-or-directory]\"\n# context: fork                       # optional\n```\n\nThe machine-readable schema for this frontmatter lives at\n[`schemas/skill.schema.json`](schemas/skill.schema.json). Validate all skills\nand role bundles locally with:\n\n```bash\nruby scripts/validate_skill_schema.rb\n```\n\nValidate skill fixture manifests and expected evidence strings with:\n\n```bash\nruby scripts/test_skill_fixtures.rb\n```\n\nValidate remediation regression manifests and expected auto-fix diffs with:\n\n```bash\nruby scripts/test_remediation_fixtures.rb\n```\n\nRegenerate the deterministic skill quality scorecard with:\n\n```bash\nruby scripts/generate_quality_scorecard.rb\nruby scripts/generate_quality_scorecard.rb --check\n```\n\nValidate framework provenance, versions, owners, and review dates with:\n\n```bash\nruby scripts/validate_framework_registry.rb\nruby scripts/validate_framework_registry.rb --stale --max-age-days 365\n```\n\nValidate required domain CODEOWNERS review gates with:\n\n```bash\nruby scripts/validate_codeowners.rb\n```\n\nRelease archives include SHA-256 checksums generated by the release workflow.\nSee [`docs/release-integrity.md`](docs/release-integrity.md) for verification\nsteps.\n\nCI/CD examples for GitHub Actions, GitLab CI, Azure DevOps, Jenkins,\npre-commit, and local agent usage are available in\n[`docs/ci-cd-examples.md`](docs/ci-cd-examples.md). Validate those examples\nlocally with:\n\n```bash\nruby scripts/validate_ci_cd_examples.rb\n```\n\n### Normalized finding JSON\n\nEvery skill must be able to emit findings as normalized JSON that validates\nagainst [`schemas/finding.schema.json`](schemas/finding.schema.json). The\ntop-level envelope, required run/skill metadata, finding fields, evidence,\nframework/CWE references, remediation fields, and test strategy requirements are\ndocumented in [`docs/normalized-json-output.md`](docs/normalized-json-output.md).\nWhen SARIF is requested, skills should map those normalized findings to\nSARIF 2.1.0-compatible JSON using [`docs/sarif-output.md`](docs/sarif-output.md).\nWhen tracker handoff is requested, skills should map normalized findings to\ntracker-ready work items using\n[`docs/tracker-handoff.md`](docs/tracker-handoff.md) and\n[`schemas/tracker-handoff.schema.json`](schemas/tracker-handoff.schema.json).\n\n### Progressive disclosure (keep `SKILL.md` lean)\n\nClaude's skill guidance: when a `SKILL.md` would exceed ~500 lines, **don't inline everything** — split detail into sibling reference files in the same directory and link to them from `SKILL.md`. The agent loads a reference only when it needs it, so the entrypoint stays cheap to load.\n\n```\nskills/appsec/threat-modeling/\n├── SKILL.md                  ← entrypoint (lean): when-to-use, rules, output format\n├── threat-actor-profiles.md  ← reference, loaded on demand\n└── csharp-dotnet.md          ← language-specific reference\n```\n\nThis is why some skills ship extra `.md` files alongside `SKILL.md` (e.g. `cloud/aws-review/benchmark-checklist.md`, `compliance/soc2-gap/tsc-criteria.md`) — it is the intended pattern, not duplication.\n\n---\n\n---\n\n## Skills\n\n45 skills across 10 security domains.\n\n### Application Security\n\n| Skill | Path | Frameworks |\n|-------|------|------------|\n| Threat Modeling (STRIDE) | `skills/appsec/threat-modeling/` | STRIDE, PASTA, MITRE ATT&CK |\n| Secure Code Review | `skills/appsec/secure-code-review/` | OWASP ASVS 4.0.3, CWE Top 25 |\n| OWASP Top 10 (Web) | `skills/appsec/owasp-top-10-web/` | OWASP Top 10 2021 |\n| API Security Review | `skills/appsec/api-security/` | OWASP API Security Top 10 2023 |\n| Dependency Scanning | `skills/appsec/dependency-scanning/` | SLSA v1.0, CycloneDX, SPDX |\n\n### AI Security\n\n| Skill | Path | Frameworks |\n|-------|------|------------|\n| LLM Top 10 Review | `skills/ai-security/llm-top-10/` | OWASP LLM Top 10 2025 |\n| Agentic AI Top 10 | `skills/ai-security/agentic-top-10/` | OWASP Agentic AI, MITRE ATLAS |\n| Prompt Injection Testing | `skills/ai-security/prompt-injection/` | OWASP LLM01:2025, MITRE ATLAS |\n| Model Supply Chain | `skills/ai-security/model-supply-chain/` | OWASP LLM03:2025, SLSA v1.0 |\n| AI Data Privacy | `skills/ai-security/ai-data-privacy/` | NIST AI RMF, OWASP LLM02:2025 |\n| Agent Security Architecture | `skills/ai-security/agent-security/` | OWASP Agentic AI, NIST AI RMF |\n\n### Identity & Access\n\n| Skill | Path | Frameworks |\n|-------|------|------------|\n| IAM Security Review | `skills/identity/iam-review/` | NIST SP 800-63B, CIS Controls v8 |\n| Access Review | `skills/identity/access-review/` | CIS Controls v8, NIST SP 800-53 |\n| RBAC/ABAC Design | `skills/identity/rbac-design/` | NIST RBAC, NIST SP 800-162 |\n| Zero Trust Assessment | `skills/identity/zero-trust-assessment/` | NIST SP 800-207, CISA ZTMM v2 |\n| Privileged Access Management | `skills/identity/privileged-access/` | CIS Controls v8, NIST SP 800-53 |\n\n### Cloud Security\n\n| Skill | Path | Frameworks |\n|-------|------|------------|\n| AWS Security Review | `skills/cloud/aws-review/` | CIS AWS Benchmark v3.0 |\n| Azure Security Review | `skills/cloud/azure-review/` | CIS Azure Benchmark v2.1 |\n| GCP Security Review | `skills/cloud/gcp-review/` | CIS GCP Benchmark v2.0 |\n| IaC Security | `skills/cloud/iac-security/` | OWASP IaC Security, SLSA v1.0 |\n| Container Security | `skills/cloud/container-security/` | CIS Docker v1.6, CIS K8s v1.9 |\n\n### Vulnerability Management\n\n| Skill | Path | Frameworks |\n|-------|------|------------|\n| CVE Triage | `skills/vuln-management/cve-triage/` | CVSS 4.0, SSVC 2.1, CISA KEV, EPSS |\n| Patch Prioritization | `skills/vuln-management/patch-prioritization/` | SSVC 2.1, EPSS, CISA KEV |\n| SBOM Analysis | `skills/vuln-management/sbom-analysis/` | CycloneDX, SPDX, VEX |\n| Scanner Tuning | `skills/vuln-management/scanner-tuning/` | CVSS 4.0, CWE |\n\n### Compliance\n\n| Skill | Path | Frameworks |\n|-------|------|------------|\n| SOC 2 Gap Analysis | `skills/compliance/soc2-gap/` | AICPA TSC |\n| ISO 27001 Gap Analysis | `skills/compliance/iso27001-gap/` | ISO 27001:2022 |\n| PCI DSS Review | `skills/compliance/pci-dss-review/` | PCI DSS v4.0 |\n| HIPAA Review | `skills/compliance/hipaa-review/` | HIPAA Security Rule |\n| NIST CSF Assessment | `skills/compliance/nist-csf-assessment/` | NIST CSF 2.0 |\n\n### Incident Response\n\n| Skill | Path | Frameworks |\n|-------|------|------------|\n| IR Playbook | `skills/incident-response/ir-playbook/` | NIST SP 800-61 |\n| Forensics Checklist | `skills/incident-response/forensics-checklist/` | NIST SP 800-86, RFC 3227 |\n| Containment Strategies | `skills/incident-response/containment/` | NIST SP 800-61, MITRE ATT&CK |\n| Post-Incident Review | `skills/incident-response/post-incident-review/` | NIST SP 800-61 |\n\n### SecOps\n\n| Skill | Path | Frameworks |\n|-------|------|------------|\n| Detection Engineering | `skills/secops/detection-engineering/` | MITRE ATT&CK v16, Sigma |\n| SIEM Rules | `skills/secops/siem-rules/` | MITRE ATT&CK v16 |\n| Alert Triage | `skills/secops/alert-triage/` | MITRE ATT&CK v16 |\n| Log Analysis | `skills/secops/log-analysis/` | MITRE ATT&CK v16, NIST SP 800-92 |\n\n### Network Security\n\n| Skill | Path | Frameworks |\n|-------|------|------------|\n| Firewall Rule Audit | `skills/network/firewall-review/` | CIS Controls v8, NIST SP 800-41 |\n| Network Segmentation | `skills/network/segmentation/` | NIST SP 800-207, CIS Controls v8 |\n| DNS Security | `skills/network/dns-security/` | NIST SP 800-81, CIS Controls v8 |\n\n### DevSecOps\n\n| Skill | Path | Frameworks |\n|-------|------|------------|\n| Pipeline Security | `skills/devsecops/pipeline-security/` | SLSA v1.0, OWASP CI/CD Top 10 |\n| Secrets Management | `skills/devsecops/secrets-management/` | OWASP Secrets Mgmt, NIST SP 800-57 |\n| SAST Configuration | `skills/devsecops/sast-config/` | OWASP ASVS, CWE Top 25 |\n| DAST Configuration | `skills/devsecops/dast-config/` | OWASP Top 10, OWASP Testing Guide |\n\n---\n\n## Role Bundles\n\nPre-configured skill sequences for common security roles. Each bundle orchestrates skills in the right order for the engagement type.\n\n| Role | Description | Skills |\n|------|-------------|--------|\n| **vCISO** | Security program leadership, risk assessment, compliance, board reporting | nist-csf-assessment, soc2-gap, iam-review, cve-triage, threat-modeling |\n| **SOC Analyst** | Alert triage, threat hunting, incident investigation, detection engineering | alert-triage, detection-engineering, ir-playbook, log-analysis, cve-triage |\n| **Security Engineer** | Building security into products and infrastructure | secure-code-review, dependency-scanning, cve-triage, secrets-management, pipeline-security, container-security, iam-review |\n| **AppSec Engineer** | Application security design, testing, and code review | threat-modeling, secure-code-review, api-security, dependency-scanning, prompt-injection, owasp-top-10-web |\n| **Cloud Security Engineer** | Cloud posture, IaC review, container security, identity | aws-review, azure-review, gcp-review, iac-security, container-security, zero-trust-assessment, privileged-access |\n\n---\n\n## What Makes This Different\n\n- **Framework-grounded.** Every skill cites real control IDs from OWASP, NIST, MITRE ATT&CK, or CIS. No invented controls. No hallucinated references.\n- **Consistent output format.** Structured findings with severity, CWE mapping, framework reference, evidence, remediation, and normalized JSON -- every time.\n- **AI-security skills that don't exist elsewhere.** OWASP LLM Top 10, Agentic AI security, prompt injection testing, model supply chain review.\n- **Multi-agent compatible.** Same skill file works with Claude Code, Gemini CLI, Cursor, Codex CLI, OpenClaw, and Kiro.\n- **Prompt-injection hardened.** Every skill reviewed against OWASP LLM01:2025. CI scans for injection patterns on every PR.\n- **Enterprise-ready.** Built by practitioners, not scraped from blog posts. Designed for real security programs.\n\n---\n\n## Disclaimer\n\nThese skills were built through extensive research against published security frameworks (OWASP, NIST, MITRE ATT&CK, CIS Controls) and reviewed by five specialized AI security agents:\n\n- **CISO Reviewer** — Strategic risk, compliance alignment, and program-level gaps\n- **Security Architect** — Framework accuracy, control ID verification, and design patterns\n- **Security Engineer** — Implementation correctness, tooling gaps, and operational feasibility\n- **AI Security Researcher** — LLM/agentic threat modeling, prompt injection hardening, and ATLAS coverage\n- **SOC Analyst** — Detection engineering, alert triage accuracy, and incident response workflows\n\nDespite this multi-layered review process, these skills may contain inaccuracies, outdated framework references, or gaps in coverage. **Validate all control IDs, framework versions, and remediation guidance against authoritative sources before using these skills in production security workflows.** Security frameworks evolve — always cross-reference with the latest published versions.\n\n---\n\n## Contribute\n\nContributions are welcome — skill reviews, improvements, and new skills. Read **[CONTRIBUTING.md](CONTRIBUTING.md)** for the quality bar, skill-format specification, and the review/PR checklist, and author new skills with **[SKILL_TEMPLATE.md](SKILL_TEMPLATE.md)**. Every skill must cite a real framework with verifiable control IDs.\n\n> **⏸️ Bounty program paused.** The paid bounty program is temporarily on hold. Contributions are still welcome and will be credited — we'll announce on [Discord](https://discord.gg/DKTZzfU9B) when bounties resume.\n\n## Security\n\nSee [SECURITY.md](SECURITY.md) for our prompt injection hardening policy and responsible disclosure process.\n\n## License\n\n[MIT](LICENSE)\n",
  "bytes": 15283,
  "sha": "afd712680cc97cc2e3adcca4901f7684b0ff6d495166baee060cab97aabeb373",
  "repo_slug": "unitoneai/securityskills",
  "fonte": "repo",
  "truncated": false,
  "api": "https://api.agentalog.com/api/listings/skl_unitoneai_securityskills_llm_top_10_9fd66b4b/readme"
}