Skip to content

pitimon/devsecops-ai-team

Repository files navigation

Claude Code Plugin v3.1.0 Tests MIT License

18 Agents 16 Skills 11 Tools MCP OWASP 2021+2025 10/10 488 CWEs 7 Compliance Frameworks QA 75/75

DevSecOps AI Team

Enterprise DevSecOps Plugin Skill Pack for Claude Code
Multi-Agent AI Security Team — 18 agents ทำงานร่วมกันเพื่อรักษาความปลอดภัยของซอฟต์แวร์ตลอด SDLC

Validation Security Scan Release Wiki


Table of Contents


Executive Summary

Metric Value
Project Type Claude Code Plugin Skill Pack (pure markdown/JSON/shell)
AI Agents 18 agents across 4 groups (Orchestrators, Specialists, Experts, Core Team)
Skills (Commands) 16 slash commands (/sast-scan, /dast-scan, /full-pipeline, /k8s-scan, /graphql-scan, ...)
Security Tools 11 open-source tools in Docker containers (Semgrep, ZAP, Nuclei, Grype, Trivy, Checkov, GitLeaks, Syft, TruffleHog, kube-bench, Nuclei-GraphQL)
MCP Tools 10 composable tools for programmatic integration
Compliance Frameworks 7 frameworks — OWASP Top 10 (2021+2025), NIST 800-53, MITRE ATT&CK, NCSA, PDPA, SOC 2, ISO 27001
CWE Mappings 488 total (OWASP 122 + NIST 100 + MITRE 93 + NCSA 62 + PDPA 30 + SOC 2 40 + ISO 27001 41)
OWASP Top 10 Coverage 10/10 categories — dual-version 2021+2025 mapping
Tests 1,302+ checks across 42 suites — all passing
QA Rounds 13 rounds, 75/75 latest (cumulative 1,300+ checks)
ROI 10,222% — 3,100 THB actual vs 320,000 THB equivalent (133x speed)
Version 3.1.0 (2026-03-03)

Key Highlights

  • 18 AI Agents, 1 Team — Orchestrators delegate งานให้ Specialists, Experts วิเคราะห์ผลข้ามเครื่องมือ, Core Team enforce quality gates — ทั้งหมดทำงานร่วมกันผ่าน mandatory routing table
  • 11 Security Tools, 1 Command/full-pipeline รันทุกเครื่องมือแบบ parallel, deduplicate ผลข้าม tools, สร้าง unified report ในคำสั่งเดียว
  • Real-Time Protection — บล็อก commits ที่มี CRITICAL findings, ตรวจจับ secrets (AWS keys, GitHub tokens, JWT) ก่อนเขียนลง disk — ทำงานใน 500ms
  • CVSS v4.0 Prioritization — วิเคราะห์ business impact, exploitability (Weaponized → None), กำหนด SLA tiers (P1: 24 ชม. → P4: backlog)
  • 488 CWE Compliance Mappings — Auto-map ผลสแกนไปยัง OWASP Top 10 (122), NIST 800-53 (100), MITRE ATT&CK (93), NCSA (62), PDPA (30), SOC 2 (40), ISO 27001 (41)
  • NCSA Web Security Standard — รองรับมาตรฐานความมั่นคงปลอดภัยเว็บไซต์ สพธอ. (HTTP Headers, TLS, Session Management)
  • MCP Server — 10 composable tools สำหรับ programmatic integration กับ MCP-compatible clients (compare, compliance_status, suggest_fix, history, pipeline)
  • Custom OWASP Rules — 84 custom Semgrep rules for A01 (access control), A02 (crypto), A03 (injection), A04 (insecure design), A05 (misconfig), A06 (vulnerable components), A07 (auth failures), A08 (integrity failures), A09 (logging), A10 (SSRF + exception handling), K8s manifests, GraphQL endpoints
  • 8 Output Formats — SARIF, JSON, Markdown, HTML, PDF, CSV, VEX, Dashboard

OWASP Top 10 Coverage

ครอบคลุม OWASP Top 10 ทั้ง 2021 และ 2025 ด้วย dual-version mapping, tools และ custom rules:

# Category (2021) Category (2025) Tools Detection Method
A01 Broken Access Control Broken Access Control Custom rules (8), ZAP, Nuclei Missing authz, IDOR, path traversal, CORS, privilege escalation (CWE-862/639/22/942/269)
A02 Cryptographic Failures Cryptographic Failures Custom rules (6), GitLeaks Weak crypto, hardcoded keys, insecure algorithms, missing encryption (CWE-327/328/330/338/916)
A03 Injection Injection Custom rules (11), ZAP, Nuclei SQLi, command injection, XSS, LDAP injection, template injection (CWE-89/78/79/90/1336)
A04 Insecure Design Insecure Design Custom rules (4), Checkov Missing rate limiting, insecure file upload, business logic bypass (CWE-770/434/840/841)
A05 Security Misconfiguration Security Misconfiguration Custom rules (6), Trivy, Checkov Debug mode, directory listing, default creds, verbose errors (CWE-16/215/548/756/1004)
A06 Vulnerable Components Vulnerable Components Custom rules (5), Grype, Syft Outdated dependencies, known CVEs, version pinning, SBOM analysis (CWE-1104/937/1035)
A07 Auth Failures Identification Failures Custom rules (5), ZAP, Nuclei Weak passwords, missing MFA, session fixation, credential stuffing (CWE-287/306/384/640/521)
A08 Data Integrity Failures Data Integrity Failures Custom rules (5), Trivy Unsafe deserialization, unsigned artifacts, CI/CD integrity (CWE-502/829/494/915/345)
A09 Logging Failures Logging Failures Custom rules (7) Missing auth logs, silent catch, PII in logs, log injection, rate-limit logging
A10 SSRF Exception Handling (new 2025) Custom rules (7+4), ZAP, Nuclei SSRF: cloud metadata, DNS rebinding + Exception: stack trace leak, global catch (CWE-918/209/392)

What's New

ดู CHANGELOG.md สำหรับรายละเอียดทั้งหมด (v1.0.0 → v3.1.0)

v3.1.0 — Commercial Ready (Latest)

  • Professional README Redesign — Commercial-grade product presentation (1,071 → 615 lines)
  • Quick Start Guide — Install to first scan in 5 minutes (docs/QUICK-START.md)
  • First Scan Walkthrough — Behind-the-scenes agent orchestration explained (docs/FIRST-SCAN-WALKTHROUGH.md)
  • Demo Scenarios — 3 presenter-ready demos (5/10/15 min) with bilingual talk track (demo/)
  • Demo Vulnerable Project — Onboarding sample with intentional vulnerabilities (tests/fixtures/demo-project/)
  • Service Tiers — Starter / Pro / Enterprise consulting + training offerings
  • Consolidated Documentation — Architecture, Features, Project Structure extracted to dedicated docs
Previous versions

v3.0.4 — Platform Release

  • SQLite Historical Databasescripts/scan-db.sh with 7 subcommands for persistent scan tracking and trend analysis
  • DAG Pipeline Enginerunner/pipeline-engine.sh with topological sort and 4 built-in pipeline definitions
  • Security Dashboard — Alpine.js + Chart.js self-contained HTML dashboard with 6 panels
  • K8s Security Scanning/k8s-scan skill with 8 Semgrep rules + kube-bench CIS Benchmark integration
  • GraphQL Security Scanning/graphql-scan skill with 8 Semgrep rules + 4 Nuclei templates
  • 2 New MCP Toolsdevsecops_history and devsecops_pipeline (10 total)
  • 84 Custom Semgrep Rules — +16 new rules (8 K8s + 8 GraphQL)

v2.8.0 — Supply Chain Compliance + OWASP 10/10

  • OWASP 10/10 Custom Rules — A06 Vulnerable Components (5), A07 Auth Failures (5), A08 Integrity Failures (5) — total 68 custom Semgrep rules
  • SLSA Provenance Assessment/slsa-assess skill with SLSA v1.1 reference for EU CRA compliance
  • VEX Output Format — CycloneDX VEX + OpenVEX as 7th output format
  • TruffleHog Secret Scanning — 9th security tool with git/filesystem/s3 modes
  • SOC 2 + ISO 27001 Compliance — 2 new mapping files (~81 CWEs total), 7 frameworks

v2.7.0 — OWASP 2025, Nuclei DAST, PDPA Compliance

  • OWASP Top 10 2025 — dual-version mapping (2021+2025) across all 122 CWEs, 4 new rule sets (A02/A04/A05/A10-exception)
  • 53 Custom Semgrep Rules — was 33, added A02 Cryptographic Failures (6), A04 Insecure Design (4), A05 Misconfiguration (6), A10 Exception Handling (4)
  • Nuclei DAST Integration — second DAST tool alongside ZAP with Docker, dispatcher, normalizer
  • PDPA Compliance — 30 CWE mappings for Thai Personal Data Protection Act
  • NCSA 1.0 Enhanced — Permissions-Policy, COOP, COEP, TLS 1.3 checks

v2.5.0 — Custom OWASP Rules (A01/A03/A10), 3 New MCP Tools, PDF/CSV

  • A01/A03/A10 Custom Semgrep Rules — 26 new rules ตรวจจับ access control, injection, SSRF anti-patterns
  • 3 New MCP Toolsdevsecops_compare (trend diff), devsecops_compliance_status (aggregate compliance), devsecops_suggest_fix (remediation)
  • PDF/CSV Formatters — enterprise PDF export (pandoc), spreadsheet CSV export
  • 700+ tests across 19 suites (was 587)

v2.4.0 — DAST Infrastructure, A09 Detection & NCSA Validation

  • A09 Custom Semgrep Rules — 7 rules (5 categories) ตรวจจับ OWASP A09:2021 anti-patterns (CWE-117/390/532/778) ใน Python + JS/TS
  • ZAP Multi-Mode Dispatcher — 3 modes: baseline (passive, 120s), full (active, 1800s), api (OpenAPI, 600s) + authenticated scanning
  • NCSA Website Security Validator — ตรวจ HTTP Security Headers (1.x), Transport Security (2.x), Session Management (4.x) ตามมาตรฐาน สพธอ.
  • DAST Live Testing — conditional test suite สำหรับ live ZAP scan (ต้องตั้ง DAST_TARGET)
  • 587 tests across 15 suites (was 461)

v2.3.0 — NCSA Compliance Mapping

  • NCSA Website Security Standard v1.0 — 62 CWE mappings across 7 categories (มาตรฐานความมั่นคงปลอดภัยเว็บไซต์ สพธอ.)
  • MCP ncsa framework supportdevsecops_compliance รองรับ NCSA framework
  • Auto-Fix Skill/auto-fix อ่านผล scan → สร้าง patch → ขออนุมัติ → แก้ code → re-scan

v2.2.0 — Framework Remediation + Test Hardening

  • Framework-Aware Remediation — 4 reference files (Django, React/Next.js, Express/Node, Spring) + auto-detection routing
  • Syft Normalizer — SBOM component inventory from CycloneDX-JSON
  • 65 new functional tests — hooks (27), dedup (15), MCP handlers (23)

v2.1.0 — Security Fixes + Coverage Gaps

  • Security Fixes — Python3 dependency guard, MCP command injection fix (execFile), ZAP OOM memory limits
  • RBAC Gate — role-based policy (developer/security-lead/release-manager)
  • Zod Validation — MCP inputs validated ด้วย Zod schemas ทั้ง 5 tools

v2.0.0 — MCP Server + Agent Orchestration

  • MCP Server — 5 MCP tools สำหรับ programmatic integration
  • Agent Orchestration — mandatory routing table + delegation chain
  • Cross-tool Dedup — deduplicate ด้วย (cve_id, file, line)
  • Smart Detection — session-start.sh ตรวจจับ tech stack อัตโนมัติ

v1.0.0 — Initial Release

  • 18 AI agents, 12 skills, 7 Docker security tools
  • SARIF/JSON/Markdown output formatters
  • CWE → OWASP/NIST/MITRE compliance mappings

Use Cases

1. Development Team — Daily Security Scanning

Developer เปิด Claude Code ในโปรเจค → session-start ตรวจจับ tech stack อัตโนมัติ
→ แก้ไขโค้ด → scan-on-write ตรวจจับ secrets ทันที (500ms)
→ พิมพ์ /sast-scan → พบ SQL Injection → /auto-fix สร้าง patch + re-scan
→ git commit → pre-commit-gate บล็อกถ้ามี CRITICAL ค้างอยู่

2. Security Team — Pre-Release Assessment

Security Lead เรียก /full-pipeline → รัน 7 tools parallel
→ vuln-triager จัดลำดับตาม CVSS v4.0 + exploitability
→ compliance-officer map ไปยัง OWASP/NIST/MITRE/NCSA
→ /security-gate ตัดสินใจ PASS/FAIL ตาม role-based policy
→ report-generator สร้าง HTML dashboard + SARIF สำหรับ GitHub Security tab

3. Compliance Officer — Audit & Reporting

เรียก /compliance-report --framework all
→ ได้ cross-walk matrix: CWE → OWASP Top 10 + NIST 800-53 + MITRE ATT&CK + NCSA
→ export เป็น SARIF upload ไป GitHub / JSON ส่งเข้า SIEM
→ /incident-response สร้าง IR playbook ตาม NIST 800-61 เมื่อพบ CRITICAL

4. CI/CD Pipeline — Automated Gate

# GitHub Actions example
- name: Security Gate
  run: |
    # MCP tool: scan → gate → report
    devsecops_scan --tool semgrep --target .
    devsecops_scan --tool grype --target .
    devsecops_gate --policy security-lead

5. DAST — Web Application Testing

Security Engineer ตั้งค่า DAST_TARGET → เลือก scan mode:
→ baseline (CI/CD, passive only, 120s)
→ full (pre-release, active attacks, 1800s)
→ api (OpenAPI spec-driven, 600s)
→ NCSA validator ตรวจ HTTP headers + TLS + session cookies
→ ผลรวมกับ SAST findings ใน unified report

Why DevSecOps AI Team?

ปัญหา: ทีม Dev ต้องใช้เครื่องมือ security หลายตัว แต่ละตัวมี CLI, output format, และวิธีตีความผลต่างกัน ทำให้เสียเวลาในการเรียนรู้และ integrate เข้ากับ workflow

วิธีแก้: Plugin นี้รวม 7 เครื่องมือ security ชั้นนำ (ทั้งหมด Open Source) ให้ทำงานผ่าน Claude Code ด้วยภาษาธรรมชาติ — พิมพ์ /sast-scan แทนการจำ CLI ยาวๆ ได้ผลลัพธ์ที่เข้าใจง่ายพร้อมคำแนะนำการแก้ไข

ก่อนใช้ Plugin หลังใช้ Plugin
semgrep scan --config p/owasp-top-ten --sarif ... /sast-scan
gitleaks detect --source . --report-format json ... /secret-scan
trivy image --severity HIGH,CRITICAL myapp:latest ... /container-scan
ต้องรัน 7 tools แยกกัน แล้ว merge ผลเอง /full-pipeline (รันทั้ง 7 tools พร้อมกัน)
เปิด spreadsheet map CWE → NIST เอง /compliance-report --framework all
ถกกันว่า deploy ได้ไหม /security-gate (ตัดสินอัตโนมัติตาม policy)

Quick Start

1. Install Plugin

# Step 1: ลงทะเบียน marketplace
claude plugin marketplace add pitimon/devsecops-ai-team

# Step 2: ติดตั้ง plugin
claude plugin install devsecops-ai-team@pitimon-devsecops

ทางเลือก: ติดตั้งจาก local directory

git clone https://github.com/pitimon/devsecops-ai-team.git
claude plugin marketplace add ./devsecops-ai-team
claude plugin install devsecops-ai-team@pitimon-devsecops

2. ตรวจสอบ Prerequisites

# ต้องมี Docker Engine 20.10+ และ Docker Compose v2+
bash scripts/check-prerequisites.sh

3. (Optional) ติดตั้ง MCP Server

cd mcp && npm install

MCP server จะถูก load อัตโนมัติผ่าน .mcp.json เมื่อเปิด Claude Code session

4. เริ่มใช้งาน

# เปิด Claude Code แล้วพิมพ์
/devsecops-setup          # ตรวจจับ tech stack + ตั้งค่าอัตโนมัติ
/secret-scan              # สแกนหา secrets ที่หลุดเข้า codebase
/sast-scan                # สแกนช่องโหว่ในโค้ด (SQL Injection, XSS, ...)
/full-pipeline            # รันทุก scan แบบ parallel แล้วสรุปผล

หมายเหตุ: ทุกเครื่องมือทำงานใน Docker containers บนเครื่อง local — ไม่ส่ง source code ไปที่ไหน

Full installation guide: docs/INSTALL.md | Quick start walkthrough: docs/QUICK-START.md


Architecture Overview

      You (Claude Code)
           |
           +--- Skill commands (/sast-scan, /full-pipeline, ...)
           |
           +--- MCP tools (devsecops_scan, devsecops_gate, ...)   <-- v2.0
           |
           v
+------------------------------------------------------------------+
|                      18 AI Agents                                 |
|                                                                   |
|  +---------------+  +---------------+  +------------------------+ |
|  | Orchestrators |  |  Specialists  |  |  Experts + Core Team   | |
|  |  (3 agents)   |  |  (7 agents)   |  |     (8 agents)         | |
|  |               |  |               |  |                        | |
|  | devsecops-    |  | sast          |  | compliance-officer     | |
|  |   lead <------+--+ dast          |  | threat-modeler         | |
|  |   (router)    |  | sca           |  | vuln-triager           | |
|  | stack-        |  | container     |  | remediation-advisor    | |
|  |   analyst     |  | iac           |  | code-reviewer          | |
|  | team-         |  | secret        |  | incident-responder     | |
|  |   configurator|  | sbom          |  | report-generator       | |
|  |               |  |               |  | pipeline-guardian      | |
|  +---------------+  +---------------+  +------------------------+ |
+----------------------------+--------------------------------------+
                             | bash -> job-dispatcher.sh
                             v
+------------------------------------------------------------------+
|              Sidecar Runner (Alpine + Docker CLI)                 |
|      job-dispatcher.sh -> result-collector.sh -> normalize        |
|                  -> dedup-findings.sh -> format                   |
+--+------+------+------+------+------+-------+-------+------------+
   |      |      |      |      |      |       |       |
 +-v-+ +--v--++--v--++--v--++--v--++--v--++--v--++---v---+
 |Sem| |Grype||Trivy||Chek ||GitL || ZAP ||Syft ||Truf  |
 |gre| |     ||     ||ov   ||eaks ||     ||     ||fleHog|
 |p  | | SCA || Con || IaC || Sec ||DAST ||SBOM || Sec  |
 +---+ +-----++-----++-----++-----++-----++-----++------+
         All tools run locally in Docker containers

How It Works

  1. คุณพิมพ์คำสั่ง เช่น /sast-scan ใน Claude Code (หรือเรียกผ่าน MCP tool)
  2. Orchestrator (devsecops-lead) วิเคราะห์ request แล้ว MUST delegate ไปยัง specialist ตาม routing table
  3. Specialist agent ส่งงานผ่าน job-dispatcher.sh ไปยัง Docker container
  4. Tool (เช่น Semgrep) รันใน container แล้วส่งผลกลับ
  5. json-normalizer.sh แปลงผลเป็น Unified Finding Schema (severity mapped ถูกต้อง)
  6. dedup-findings.sh รวมผลจากหลาย tools แล้วตัด duplicate ออก
  7. Expert agents วิเคราะห์: จัดลำดับความสำคัญ, map compliance, แนะนำการแก้ไข
  8. Report generator สร้างรายงานในรูปแบบที่ต้องการ (SARIF/JSON/Markdown/HTML)

See Architecture Reference for pipeline delegation and decision loop details


Features at a Glance

Category Details
16 Skills 10 scanning + 6 orchestration slash commands — /sast-scan, /dast-scan, /full-pipeline, /k8s-scan, /graphql-scan, and more. See Feature Reference
18 AI Agents 3 Orchestrators, 7 Specialists, 4 Experts, 4 Core Team — mandatory routing + delegation chain. See Agent Catalog
Output Formats SARIF, JSON, Markdown, HTML, PDF, CSV, VEX, Dashboard
10 MCP Tools devsecops_scan, devsecops_results, devsecops_gate, devsecops_compliance, devsecops_status, devsecops_compare, devsecops_compliance_status, devsecops_suggest_fix, devsecops_history, devsecops_pipeline. See MCP Reference
7 Compliance Frameworks OWASP Top 10 (2021+2025), NIST 800-53, MITRE ATT&CK, NCSA, PDPA, SOC 2, ISO 27001 — 488 CWE mappings total
3 Autonomous Hooks session-start (tech stack detection), scan-on-write (real-time secret/injection detection), pre-commit-gate (block CRITICAL findings)
Sidecar Runner Docker container orchestration — minimal mode (oneshot) or full mode (persistent). See Runner Reference
84 Custom Rules Semgrep rules covering OWASP A01-A10 + K8s manifests + GraphQL endpoints
CVSS v4.0 Triage Exploitability classification (Weaponized/Active/POC/Theoretical/None) + SLA priority matrix (P1: 24h → P4: backlog)
RBAC Security Gate Role-based policy (developer/security-lead/release-manager) with configurable severity thresholds

Security & Privacy

  • Source code ไม่ออกจากเครื่อง — ทุก tool รันใน local Docker containers
  • ผลสแกนอยู่ใน RAM — ใช้ tmpfs volume, ไม่เขียนลง disk
  • Workspace mount เป็น read-only — tools อ่านได้อย่างเดียว แก้ไขไม่ได้
  • ไม่มี network access — containers ไม่ต้องการ internet (ยกเว้น ZAP ที่ต้องเข้าถึง target URL)
  • Non-root containers — Dockerfile ใช้ USER ที่ไม่ใช่ root + tini init

ดู SECURITY.md สำหรับ vulnerability reporting policy


Testing & Quality

Test Results (1,302+)

Validation:          276/276 structural checks (plugin structure, skills, agents, mappings)
Normalizer:           41/41  severity mapping + multi-array + null safety
MCP Server:           30/30  config + syntax + tool definitions
MCP Handlers:         37/37  Zod validation + gate logic + compliance crosswalk + NCSA + PDPA
Hooks:                27/27  session-start + scan-on-write + pre-commit-gate
Dedup:                15/15  cross-tool deduplication
Auto-Fix:             37/37  SKILL.md structure + agent config + routing
DAST Integration:     22/22  ZAP fixture + normalizer + dispatcher
MCP Integration:      38/38  Docker availability + handler logic + runner
A01 Rules:            30/30  A01 access control rule YAML + metadata + CWE + OWASP 2025
A02 Rules:            17/17  A02 cryptographic failures rule YAML + metadata + CWE
A03 Rules:            33/33  A03 injection rule YAML + metadata + CWE + OWASP 2025
A04 Rules:            17/17  A04 insecure design rule YAML + metadata + CWE
A05 Rules:            18/18  A05 security misconfiguration rule YAML + metadata + CWE
A06 Rules:            17/17  A06 vulnerable components rule YAML + metadata + CWE
A07 Rules:            18/18  A07 authentication failures rule YAML + metadata + CWE
A08 Rules:            16/16  A08 integrity failures rule YAML + metadata + CWE
A09 Rules:            28/28  A09 logging rule YAML + metadata + CWE + OWASP 2025
A10 Rules:            31/31  A10 SSRF + exception handling rules + OWASP 2025
ZAP Modes:            34/34  mode parsing + timeout + Docker commands + fixtures
NCSA Validator:       28/28  script structure + header checks + TLS 1.3 + output format
MCP Compare:          22/22  compare + compliance_status + suggest_fix tools
DAST Live:             0/0   conditional (requires DAST_TARGET env var)
Nuclei Integration:   22/22  Nuclei fixture + normalizer + dispatcher
PDPA Mapping:         17/17  PDPA CWE mappings + structure + coverage
SOC 2 Mapping:        17/17  SOC 2 Trust Service Criteria mappings
ISO 27001 Mapping:    17/17  ISO 27001 Annex A control mappings
SLSA Skill:           15/15  SLSA provenance assessment skill
VEX Formatter:        20/20  CycloneDX + OpenVEX output format
TruffleHog:           21/21  TruffleHog fixture + normalizer + dispatcher
Secret Verifier:      18/18  secret validity checking + provider verification
Formatters:           29/29  SARIF + JSON + Markdown + HTML + CSV + PDF formatter validation
Runner:               28/28  job-dispatcher + result-collector + Docker orchestration
Version Bump:         17/17  version-bump.sh script tests
CI Adapter:           25/25  CI platform detection + adapter functions
CI Templates:         65/65  GitHub Actions + GitLab CI template validation
Release:              12/12  release checklist script tests
Scan DB:              39/39  SQLite scan history database + 7 subcommands + OWASP enrichment + compliance + multi-tool lifecycle
Pipeline Engine:      25/25  DAG pipeline engine + topological sort + cycle detection
K8s Scan:             23/23  K8s skill + rules + kube-bench + normalizer integration
GraphQL Scan:         34/34  GraphQL skill + rules + Nuclei templates + normalizer + metadata
Dashboard:            26/26  dashboard generator + template + data injection + special character regression
--------------------------------------------------------------
Total:              1302+ checks passed (42 suites)

ROI & Business Value

ข้อมูลจาก MANDAY-ESTIMATION.md — การวิเคราะห์ ROI เปรียบเทียบกับการพัฒนาแบบ manual

Development Cost Comparison

Metric Manual Development Claude Code (Actual) Savings
Duration 40 man-days (2 months) 2.4 days (~19 hours) 133x faster
Cost 320,000 THB 3,100 THB 99% less
ROI 10,222%
Break-even 3.1 hours of manual work saved

3-Year TCO Analysis

Cost Component Traditional Team AI-Assisted Savings
Initial development 320,000 THB 3,100 THB 316,900 THB
Annual maintenance 384,000 THB/yr 6,000 THB/yr 378,000 THB/yr
3-Year Total 1,304,000 THB 93,700 THB 1,210,300 THB (92.8%)

Value Delivered

Deliverable Quantity
AI Agents 18 (fully configured with routing + delegation)
Security Skills 16 (10 scanning + 6 orchestration)
Docker Tool Integrations 11 (Semgrep, ZAP, Nuclei, Grype, Trivy, Checkov, GitLeaks, Syft, TruffleHog, kube-bench, Nuclei-GraphQL)
MCP Tools 10 (scan, results, gate, compliance, status, compare, compliance_status, suggest_fix, history, pipeline)
CWE Compliance Mappings 488 across 7 frameworks
Custom Security Rules 84 (OWASP A01-A10 + K8s + GraphQL Semgrep rules)
Automated Tests 1,302+ across 42 suites
Reference Documents 19 domain knowledge files (~500-800 lines each)
Hooks (Real-time Protection) 3 (session-start, scan-on-write, pre-commit-gate)
Output Formatters 8 (SARIF, JSON, Markdown, HTML, PDF, CSV, VEX, Dashboard)

Comparison with Alternatives

Feature DevSecOps AI Team GitHub Advanced Security Snyk SonarQube
Pricing Free (MIT) $49/user/mo $52/user/mo+ $150+/mo
SAST Semgrep CodeQL Snyk Code SonarQube
DAST ZAP + Nuclei
SCA Grype Dependabot Snyk Open Source
Container Trivy Snyk Container
IaC Checkov Snyk IaC
Secrets GitLeaks + TruffleHog Secret scanning
SBOM Syft
AI Agents 18 agents
NLP Interface Natural language
NCSA Compliance Built-in
Offline/Air-gapped Full support Partial Self-hosted
Data Privacy 100% local Cloud Cloud Self-hosted
Custom Rules 84 rules (10 OWASP + K8s + GraphQL) Custom CodeQL Custom rules
MCP Integration 10 tools

Note: DevSecOps AI Team ใช้ open-source tools ทั้งหมด — ไม่มี vendor lock-in, source code ไม่ออกจากเครื่อง


Services — บริการ Consulting + Training

DevSecOps AI Team พร้อมให้บริการ consulting, implementation, และ training สำหรับองค์กรในประเทศไทย

Tier รายละเอียด เหมาะสำหรับ
Starter Training + Workshop — สอนทีมใช้ plugin, OWASP, NCSA compliance ทีมที่เริ่มต้น DevSecOps
Pro Implementation + Training — setup pipeline, customize rules, integrate CI/CD, train ทีม องค์กรที่ต้องการ hands-on setup
Enterprise Managed Security Service — scan ต่อเนื่อง, วิเคราะห์ผล, รายงานรายเดือน, priority support องค์กรที่ต้องการ expert ดูแลต่อเนื่อง

สนใจ? ติดต่อผ่าน GitHub Issues หรือ email


Requirements

Requirement Minimum Recommended หมายเหตุ
Docker Engine 20.10+ 25.0+ จำเป็นสำหรับทุก scan
Docker Compose v2.0+ v2.24+ จำเป็นสำหรับ full mode
Node.js 18+ 20+ จำเป็นสำหรับ MCP server
Python 3.8+ 3.12+ จำเป็นสำหรับ formatters
Disk Space 2 GB 5 GB Docker images
Claude Code Latest Latest

Documentation

Document เนื้อหา
Wiki Comprehensive documentation with ASCII diagrams
QUICK-START.md Install to first scan in 5 minutes
FIRST-SCAN-WALKTHROUGH.md Behind the scenes — agent orchestration explained
INSTALL.md วิธีติดตั้งแบบ standard, manual, air-gapped, MCP
ARCHITECTURE.md Pipeline delegation, decision loop, system design
FEATURES.md Skills, agents, MCP, compliance, output formats
PROJECT-STRUCTURE.md Directory tree + file descriptions
AGENT-CATALOG.md รายละเอียด 18 agents พร้อม routing cues + triggers
TROUBLESHOOTING.md แก้ปัญหาที่พบบ่อย (14 scenarios)
FRAMEWORK-UPDATE-RUNBOOK.md ขั้นตอนอัพเดท framework versions
MANDAY-ESTIMATION.md ROI analysis + cost comparison (10,222% ROI)
CLAUDE.md Architecture + contributing guidelines
CHANGELOG.md Version history (v1.0.0 → v3.1.0)
SECURITY.md Vulnerability reporting policy

Contributing

  1. Fork the repository
  2. Create a feature branch: git checkout -b feat/my-feature
  3. Follow Conventional Commits: feat:, fix:, docs:, etc.
  4. Run validation: bash tests/validate-plugin.sh
  5. Run all tests: for f in tests/validate-plugin.sh tests/test-*.sh; do bash "$f"; done
  6. Submit a Pull Request

ดูรายละเอียดใน CLAUDE.md


Roadmap

Version Status Theme Key Features
v1.0.0 Released Foundation 18 agents, 12 skills, 7 tools, compliance mappings
v2.0.0 Released MCP & Quality MCP server, orchestration, dedup, smart detection
v2.1.0 Released Security & RBAC 3 security fixes, RBAC gate, Zod validation
v2.2.0 Released Remediation Framework-aware fixes (Django/React/Express/Spring)
v2.3.0 Released Automation /auto-fix skill, NCSA compliance, integration tests
v2.4.0 Released Detection A09 custom rules, ZAP multi-mode, NCSA validator
v2.5.0 Released Rules & MCP 33 rules (A01/A03/A09/A10), 3 MCP tools, PDF/CSV
v2.6.x Released CI/CD GitHub Actions, GitLab CI, SARIF per-tool, MCP bundle, tech debt
v2.7.0 Released OWASP 2025 Dual mapping, 53 rules, Nuclei, NCSA 1.0, PDPA
v2.8.0 Released Supply Chain SLSA, VEX, 10/10 OWASP rules, SOC 2, ISO 27001, TruffleHog
v3.0.x Released Platform SQLite DB, DAG pipeline, dashboard, K8s scan, GraphQL scan
v3.1.0 Current Commercial README redesign, onboarding, demos, service tiers

ดูรายละเอียด roadmap ทั้งหมดได้ที่ docs/PRD.md


License

MIT License — see LICENSE


DevSecOps AI Team — Enterprise Security Plugin for Claude Code
18 AI Agents | 16 Skills | 11 Tools | 10 MCP Tools | 8 Output Formats | 488 CWE Mappings | OWASP 10/10
Built with Claude Code | Powered by Open Source Security Tools

GitHub | Wiki | Releases | Issues

About

DevSecOps AI Team — Claude Code Plugin with 18 AI Agents, 13 Skills, 7 Security Tools (Semgrep, ZAP, Grype, Trivy, Checkov, GitLeaks, Syft) running in Docker containers. Full pipeline: SAST, DAST, SCA, Container, IaC, Secrets, SBOM, Compliance, IR. OWASP Top 10 10/10 coverage. 360 CWE mappings (OWASP/NIST/MITRE/NCSA).

Topics

Resources

License

Security policy

Stars

Watchers

Forks

Packages

 
 
 

Contributors