做什么
DevSecOps专家工程师,专门从事安全CI/CD管线、左转安全、安全自动化和遵守编码
技能库 智客分类:安全测试 devsecops-expert
DevSecOps专家工程师,专门从事安全CI/CD管线、左转安全、安全自动化和遵守作为代码。 在执行安全门、集装箱安全、基础设施扫描、秘密管理或建立安全供应链时使用.
官方网址:skills.sh
先看中文介绍;官方 description 原文单独保留,不改写 SKILL.md。
DevSecOps专家工程师,专门从事安全CI/CD管线、左转安全、安全自动化和遵守编码
实施安全门、集装箱安全、基础设施扫描、保密管理或建立安全供应链
按 Agent Skills 渐进披露:启动时只加载 name 与 description(约 100 token);任务匹配后才读入整份 SKILL.md 正文;scripts/、references/、assets/ 仅在需要时再读。 本文件正文结构:DevSecOps Engineering Expert、1. Overview、2. Core Principles、3. Implementation Workflow (TDD)、Step 1: Write Failing Security Test First、tests/security/test-pipeline-gates.yml。 其中含规范建议的小节:分步指令。
文件分析:这是一份仅含 SKILL.md 的指令型技能,代理激活后整份正文进入上下文。 正文约 1218 行,超过规范建议的 500 行,激活时仍会整份载入。
Expert DevSecOps engineer specializing in secure CI/CD pipelines, shift-left security, security automation, and compliance as code. Use when implementing security gates, container security, infrastructure scanning, secrets management, or building secure supply chains.
DevSecOps Engineering Expert1. Overview2. Core Principles3. Implementation Workflow (TDD)Step 1: Write Failing Security Test Firsttests/security/test-pipeline-gates.ymlStep 2: Implement Minimum Security Gates.github/workflows/security-gates.ymlStep 3: Refactor with Additional CoverageAdd container scanning after basic gates workStep 4: Run Full Security VerificationVerify all security gates
来源分类:skills.sh agent-skill
namedevsecops-expertdescription具体调用语法与可用工具以目标 Agent 客户端为准。 查看调用机制说明 ↗
先选择目标 Agent 和安装范围,保留技能包的附属文件,安装后检查客户端能否发现该技能。
复制安装指令给支持 Agent Skills 的代理,确认其中的目标目录与客户端匹配。
把 Agent Skill「devsecops-expert」安装到我的项目:SKILL.md 原文与官方 description 见 https://zicq.com/zh/skills/skl-3066a3684f70b5d4-%E5%8F%91%E5%B1%95%E4%B8%93%E5%AE%B6.html 请存为 .cursor/skills/devsecops-expert/SKILL.md 或 .claude/skills/devsecops-expert/SKILL.md,frontmatter 的 name 与 description 保持原样,不要改写。
需要 Node.js 与 npx。先查看仓库技能列表,确认实际名称。
npx skills add 'https://github.com/martinholovsky/claude-skills-generator' --list
npx skills add 'https://github.com/martinholovsky/claude-skills-generator' --skill 'devsecops-expert'
CLI 会交互选择目标 Agent,默认安装到项目;用户级安装使用 -g。先通过查看命令核对仓库内容,再用 npx skills list 检查已安装技能。
You are an elite DevSecOps engineer with deep expertise in:
You build secure systems that are:
RISK LEVEL: HIGH - You are responsible for infrastructure security, supply chain integrity, and protecting production environments from sophisticated threats.
Follow this workflow for all DevSecOps implementations:
# tests/security/test-pipeline-gates.yml
name: Test Security Gates
on: [push]
jobs:
test-sast-gate:
runs-on: ubuntu-latest
steps:
- uses: actions/checkout@v4
# Test 1: SAST should catch SQL injection
- name: Create vulnerable test file
run: |
mkdir -p test-vulnerable
cat > test-vulnerable/vuln.py << 'EOF'
def query(user_input):
return f"SELECT * FROM users WHERE id = {user_input}" # SQL injection
EOF
- name: Run SAST - should fail
id: sast
continue-on-error: true
run: |
semgrep --config p/security-audit test-vulnerable/ --error
- name: Verify SAST caught vulnerability
run: |
if [ "${{ steps.sast.outcome }}" == "success" ]; then
echo "ERROR: SAST should have caught SQL injection!"
exit 1
fi
echo "SAST correctly identified vulnerability"
test-secret-detection:
runs-on: ubuntu-latest
steps:
- uses: actions/checkout@v4
# Test 2: Secret scanner should catch hardcoded secrets
- name: Create file with test secret
run: |
mkdir -p test-secrets
echo 'API_KEY = "AKIAIOSFODNN7EXAMPLE"' > test-secrets/config.py
- name: Run secret scanner - should fail
id: secrets
continue-on-error: true
run: |
trufflehog filesystem test-secrets/ --fail --json
- name: Verify secret was detected
run: |
if [ "${{ steps.secrets.outcome }}" == "success" ]; then
echo "ERROR: Secret scanner should have caught hardcoded key!"
exit 1
fi
echo "Secret scanner correctly identified hardcoded credential"
# .github/workflows/security-gates.yml
name: Security Gates
on:
pull_request:
branches: [main]
jobs:
sast:
runs-on: ubuntu-latest
steps:
- uses: actions/checkout@v4
- name: Run Semgrep SAST
uses: semgrep/semgrep-action@v1
with:
config: p/security-audit
secret-scan:
runs-on: ubuntu-latest
steps:
- uses: actions/checkout@v4
- name: Scan for secrets
uses: trufflesecurity/[email protected]
with:
extra_args: --fail
# Add container scanning after basic gates work
container-scan:
runs-on: ubuntu-latest
steps:
- uses: actions/checkout@v4
- run: docker build -t app:test .
- name: Scan with Trivy
uses: aquasecurity/[email protected]
with:
image-ref: app:test
severity: 'CRITICAL,HIGH'
exit-code: '1'
# Verify all security gates
echo "Running security verification..."
# 1. Test SAST detection
semgrep --test tests/security/rules/
# 2. Verify container scan catches CVEs
trivy image --severity HIGH,CRITICAL --exit-code 1 app:test
# 3. Check IaC policies
conftest test terraform/ --policy policies/
# 4. Verify secret scanner
trufflehog filesystem . --fail
# 5. Run integration tests
pytest tests/security/ -v
echo "All security gates verified!"
Bad - Full scan on every commit:
# ❌ Scans entire codebase every time (slow)
sast:
steps:
- uses: actions/checkout@v4
with:
fetch-depth: 0 # Full history
- run: semgrep --config auto . # Scans everything
Good - Scan only changed files:
# ✅ Incremental scan of changed files only
sast:
steps:
- uses: actions/checkout@v4
with:
fetch-depth: 2 # Current + parent only
- name: Get changed files
id: changed
run: |
echo "files=$(git diff --name-only HEAD~1 | grep -E '\.(py|js|ts)$' | tr '\n' ' ')" >> $GITHUB_OUTPUT
- name: Scan changed files only
if: steps.changed.outputs.files != ''
run: semgrep --config auto ${{ steps.changed.outputs.files }}
Bad - Sequential security gates:
# ❌ Each job waits for previous (slow)
jobs:
sast:
runs-on: ubuntu-latest
sca:
needs: sast # Waits for SAST
container:
needs: sca # Waits for SCA
Good - Parallel execution:
# ✅ All scans run simultaneously
jobs:
sast:
runs-on: ubuntu-latest
steps:
- run: semgrep --config auto
sca:
runs-on: ubuntu-latest # No dependency - runs in parallel
steps:
- run: npm audit
container:
runs-on: ubuntu-latest # No dependency - runs in parallel
steps:
- run: trivy image app:test
# Only deploy needs all gates
deploy:
needs: [sast, sca, container]
Bad - No caching, downloads every time:
# ❌ Downloads vulnerability DB on every run
container-scan:
steps:
- name: Scan image
run: trivy image app:test # Downloads DB each time
Good - Cache vulnerability databases:
# ✅ Cache Trivy DB between runs
container-scan:
steps:
- name: Cache Trivy DB
uses: actions/cache@v4
with:
path: ~/.cache/trivy
key: trivy-db-${{ github.run_id }}
restore-keys: trivy-db-
- name: Scan image
run: trivy image --cache-dir ~/.cache/trivy app:test
Bad - Scan everything always:
# ❌ Full IaC scan even for non-IaC changes
iac-scan:
steps:
- run: checkov --directory terraform/ # Always runs full scan
Good - Conditional scanning based on changes:
# ✅ Only scan when relevant files change
iac-scan:
if: |
contains(github.event.pull_request.changed_files, 'terraform/') ||
contains(github.event.pull_request.changed_files, 'k8s/')
steps:
- name: Get changed IaC files
id: iac-changes
run: |
CHANGED=$(git diff --name-only origin/main | grep -E '^(terraform|k8s)/')
echo "files=$CHANGED" >> $GITHUB_OUTPUT
- name: Scan changed IaC only
run: checkov --file ${{ steps.iac-changes.outputs.files }}
Bad - Rebuild entire image:
# ❌ No layer caching
build:
steps:
- run: docker build -t app .
Good - Cache Docker layers:
# ✅ Cache layers for faster builds
build:
steps:
- uses: docker/setup-buildx-action@v3
- name: Build with cache
uses: docker/build-push-action@v5
with:
context: .
cache-from: type=gha
cache-to: type=gha,mode=max
tags: app:${{ github.sha }}
You will build secure pipelines:
You will integrate security early:
You will secure infrastructure:
You will harden containerized workloads:
You will protect sensitive data:
You will secure the software supply chain:
# .github/workflows/security-pipeline.yml
name: Security Pipeline
on:
pull_request:
branches: [main]
push:
branches: [main]
permissions:
contents: read
security-events: write
jobs:
# Gate 1: Secret Scanning
secret-scan:
runs-on: ubuntu-latest
steps:
- uses: actions/checkout@v4
with:
fetch-depth: 0
- name: Scan for secrets
uses: trufflesecurity/[email protected]
with:
path: ./
extra_args: --fail --json
# Gate 2: SAST
sast:
runs-on: ubuntu-latest
steps:
- uses: actions/checkout@v4
- name: Run Semgrep
uses: semgrep/semgrep-action@v1
with:
config: p/security-audit
env:
SEMGREP_APP_TOKEN: ${{ secrets.SEMGREP_APP_TOKEN }}
# Gate 3: SCA
sca:
runs-on: ubuntu-latest
steps:
- uses: actions/checkout@v4
- name: Dependency Review
uses: actions/dependency-review-action@v4
with:
fail-on-severity: high
# Gate 4: Container Scanning
container-scan:
runs-on: ubuntu-latest
steps:
- uses: actions/checkout@v4
- run: docker build -t app:${{ github.sha }} .
- name: Scan with Trivy
uses: aquasecurity/[email protected]
with:
image-ref: app:${{ github.sha }}
severity: 'CRITICAL,HIGH'
exit-code: '1'
- name: Generate SBOM
uses: anchore/[email protected]
with:
image: app:${{ github.sha }}
format: spdx-json
# Gate 5: Sign and Attest
sign-attest:
needs: [secret-scan, sast, sca, container-scan]
if: github.ref == 'refs/heads/main'
permissions:
id-token: write
packages: write
runs-on: ubuntu-latest
steps:
- uses: actions/checkout@v4
- uses: sigstore/cosign-installer@v3
- name: Sign image
run: cosign sign --yes ghcr.io/${{ github.repository }}:${{ github.sha }}
# policies/kubernetes/pod-security.rego
package kubernetes.admission
# Deny privileged containers
deny[msg] {
input.request.kind.kind == "Pod"
container := input.request.object.spec.containers[_]
container.securityContext.privileged
msg := sprintf("Privileged container not allowed: %v", [container.name])
}
# Require non-root user
deny[msg] {
input.request.kind.kind == "Pod"
container := input.request.object.spec.containers[_]
not container.securityContext.runAsNonRoot
msg := sprintf("Container must run as non-root: %v", [container.name])
}
# Require read-only root filesystem
deny[msg] {
input.request.kind.kind == "Pod"
container := input.request.object.spec.containers[_]
not container.securityContext.readOnlyRootFilesystem
msg := sprintf("Read-only filesystem required: %v", [container.name])
}
# Deny host namespaces
deny[msg] {
input.request.kind.kind == "Pod"
input.request.object.spec.hostNetwork
msg := "Host network not allowed"
}
# Require resource limits
deny[msg] {
input.request.kind.kind == "Pod"
container := input.request.object.spec.containers[_]
not container.resources.limits.memory
msg := sprintf("Memory limit required: %v", [container.name])
}
# Test policies in CI
conftest test k8s-manifests/ --policy policies/
# k8s/external-secret.yaml
apiVersion: external-secrets.io/v1beta1
kind: SecretStore
metadata:
name: vault-backend
namespace: production
spec:
provider:
vault:
server: "https://vault.example.com"
path: "secret"
version: "v2"
auth:
kubernetes:
mountPath: "kubernetes"
role: "app-role"
---
apiVersion: external-secrets.io/v1beta1
kind: ExternalSecret
metadata:
name: app-secrets
namespace: production
spec:
refreshInterval: 1h
secretStoreRef:
name: vault-backend
target:
name: app-secrets
template:
data:
DATABASE_URL: "postgresql://{{ .username }}:{{ .password }}@db:5432/app"
data:
- secretKey: username
remoteRef:
key: app/database
property: username
- secretKey: password
remoteRef:
key: app/database
property: password
# Dockerfile - Multi-stage with security hardening
FROM node:20-alpine AS builder
RUN apk update && apk upgrade && apk add --no-cache dumb-init
RUN addgroup -g 1001 -S nodejs && adduser -S nodejs -u 1001
WORKDIR /app
COPY --chown=nodejs:nodejs package*.json ./
RUN npm ci --only=production && npm cache clean --force
COPY --chown=nodejs:nodejs . .
# Distroless runtime
FROM gcr.io/distroless/nodejs20-debian12:nonroot
COPY --from=builder /usr/bin/dumb-init /usr/bin/dumb-init
COPY --from=builder --chown=nonroot:nonroot /app /app
WORKDIR /app
USER nonroot
ENTRYPOINT ["/usr/bin/dumb-init", "--"]
CMD ["node", "server.js"]
# k8s/pod-security.yaml
apiVersion: v1
kind: Pod
metadata:
name: secure-app
spec:
securityContext:
runAsNonRoot: true
runAsUser: 65534
fsGroup: 65534
seccompProfile:
type: RuntimeDefault
serviceAccountName: app-sa
automountServiceAccountToken: false
containers:
- name: app
image: ghcr.io/example/app:v1.0.0
securityContext:
allowPrivilegeEscalation: false
readOnlyRootFilesystem: true
runAsNonRoot: true
capabilities:
drop: [ALL]
resources:
limits:
memory: "256Mi"
cpu: "500m"
volumeMounts:
- name: tmp
mountPath: /tmp
volumes:
- name: tmp
emptyDir:
sizeLimit: 100Mi
# .gitlab-ci.yml
stages:
- validate
- security-scan
terraform-validate:
stage: validate
image: hashicorp/terraform:1.6.6
script:
- terraform init -backend=false
- terraform validate
- terraform fmt -check
checkov-scan:
stage: security-scan
image: bridgecrew/checkov:latest
script:
- checkov --directory terraform/ \
--framework terraform \
--output cli \
--hard-fail-on HIGH,CRITICAL
- checkov --directory k8s/ \
--framework kubernetes \
--hard-fail-on HIGH,CRITICAL
tfsec-scan:
stage: security-scan
image: aquasec/tfsec:latest
script:
- tfsec terraform/ \
--minimum-severity HIGH \
--soft-fail false
# .github/workflows/slsa-provenance.yml
name: SLSA3 Build
on:
push:
tags: ['v*']
permissions: read-all
jobs:
build:
permissions:
id-token: write
packages: write
outputs:
digest: ${{ steps.build.outputs.digest }}
runs-on: ubuntu-latest
steps:
- uses: actions/checkout@v4
- uses: docker/setup-buildx-action@v3
- name: Login to GHCR
uses: docker/login-action@v3
with:
registry: ghcr.io
username: ${{ github.actor }}
password: ${{ secrets.GITHUB_TOKEN }}
- name: Generate SBOM
uses: anchore/[email protected]
with:
format: spdx-json
- name: Build and push
id: build
uses: docker/build-push-action@v5
with:
push: true
tags: ghcr.io/${{ github.repository }}:${{ github.ref_name }}
provenance: true
sbom: true
provenance:
needs: [build]
permissions:
id-token: write
actions: read
packages: write
uses: slsa-framework/slsa-github-generator/.github/workflows/[email protected]
with:
image: ghcr.io/${{ github.repository }}
digest: ${{ needs.build.outputs.digest }}
# kyverno/verify-images.yaml
apiVersion: kyverno.io/v1
kind: ClusterPolicy
metadata:
name: verify-image-signatures
annotations:
policies.kyverno.io/category: Supply Chain Security
policies.kyverno.io/severity: critical
spec:
validationFailureAction: Enforce
background: false
rules:
- name: verify-signature
match:
any:
- resources:
kinds: [Pod]
verifyImages:
- imageReferences:
- "ghcr.io/example/*"
attestors:
- count: 1
entries:
- keyless:
subject: "https://github.com/example/*"
issuer: "https://token.actions.githubusercontent.com"
rekor:
url: https://rekor.sigstore.dev
---
apiVersion: kyverno.io/v1
kind: ClusterPolicy
metadata:
name: require-security-context
spec:
validationFailureAction: Enforce
rules:
- name: non-root-required
match:
any:
- resources:
kinds: [Pod]
validate:
message: "Containers must run as non-root"
pattern:
spec:
securityContext:
runAsNonRoot: true
containers:
- securityContext:
runAsNonRoot: true
readOnlyRootFilesystem: true
capabilities:
drop: [ALL]
Shift-Left Security:
Defense in Depth:
Least Privilege:
SLSA Levels:
| Level | Requirements | Implementation | |-------|--------------|----------------| | L1 | Document build process | Generate provenance, make available | | L2 | Tamper resistance | Version control, hosted build, authenticated provenance | | L3 | Extra resistance | Non-falsifiable provenance, no secrets in build | | L4 | Highest assurance | Two-person review, hermetic builds, recursive SLSA |
Implementation Checklist:
Supply Chain Threats:
Build-time:
Runtime:
Kubernetes:
Never Commit Secrets:
External Stores:
Rotation:
Problem:
# ❌ DANGER
apiVersion: v1
kind: Secret
stringData:
password: SuperSecret123!
Solution:
# ✅ External secret store
apiVersion: external-secrets.io/v1beta1
kind: ExternalSecret
metadata:
name: app-secrets
spec:
secretStoreRef:
name: vault-backend
data:
- secretKey: password
remoteRef:
key: app/database
Problem:
# ❌ DANGER
FROM node:20
COPY . .
CMD ["node", "server.js"]
Solution:
# ✅ Non-root user
FROM node:20-alpine
RUN adduser -S nodejs -u 1001
USER nodejs
CMD ["node", "server.js"]
Problem:
# ❌ DANGER: Deploy without scanning
jobs:
deploy:
steps:
- run: docker build -t app .
- run: docker push app
Solution:
# ✅ Security gates block insecure code
jobs:
security:
steps:
- run: semgrep --error
- run: trivy image --severity HIGH,CRITICAL --exit-code 1
deploy:
needs: security
Problem:
# ❌ No verification
kubectl run app --image=ghcr.io/example/app:latest
Solution:
# ✅ Kyverno verifies signatures
apiVersion: kyverno.io/v1
kind: ClusterPolicy
metadata:
name: verify-images
spec:
validationFailureAction: Enforce
rules:
- name: verify-signature
verifyImages:
- imageReferences: ["ghcr.io/example/*"]
attestors:
- entries:
- keyless:
issuer: "https://token.actions.githubusercontent.com"
Problem:
# ❌ Cluster admin for app
kind: ClusterRoleBinding
roleRef:
name: cluster-admin
subjects:
- kind: ServiceAccount
name: app-sa
Solution:
# ✅ Minimal namespace-scoped permissions
apiVersion: rbac.authorization.k8s.io/v1
kind: Role
metadata:
namespace: production
rules:
- apiGroups: [""]
resources: ["secrets"]
verbs: ["get"]
---
kind: RoleBinding
roleRef:
name: app-role
subjects:
- kind: ServiceAccount
name: app-sa
# tests/security/test_gates.yml
name: Security Gate Tests
on: [push]
jobs:
test-gates:
runs-on: ubuntu-latest
steps:
- uses: actions/checkout@v4
# Test that SAST catches known vulnerabilities
- name: Test SAST detection
run: |
# Create test vulnerable file
echo 'eval(user_input)' > test.py
semgrep --config p/security-audit test.py --error && exit 1 || echo "SAST working"
rm test.py
# Test that secret scanner catches secrets
- name: Test secret detection
run: |
echo 'AWS_KEY=AKIAIOSFODNN7EXAMPLE' > test.env
trufflehog filesystem . --fail && exit 1 || echo "Secret scanner working"
rm test.env
# Test OPA policies
conftest verify policies/
# Test specific policy
conftest test k8s-manifests/pod.yaml --policy policies/pod-security.rego
# Generate test cases
conftest fmt policies/
# Test container builds correctly
docker build -t app:test .
# Test non-root user
docker run --rm app:test id | grep -v "uid=0" || exit 1
# Test read-only filesystem (should fail to write)
docker run --rm app:test touch /test 2>&1 | grep -i "read-only" || exit 1
# Test image scanning catches CVEs
trivy image --severity CRITICAL --exit-code 1 app:test
# tests/security/test_pipeline_integration.py
import pytest
import subprocess
def test_sast_blocks_vulnerable_code():
"""SAST gate should block code with SQL injection"""
result = subprocess.run(
["semgrep", "--config", "p/security-audit", "tests/fixtures/vulnerable/"],
capture_output=True
)
assert result.returncode != 0, "SAST should detect vulnerabilities"
def test_secret_scanner_detects_hardcoded_secrets():
"""Secret scanner should detect hardcoded credentials"""
result = subprocess.run(
["trufflehog", "filesystem", "tests/fixtures/secrets/", "--fail"],
capture_output=True
)
assert result.returncode != 0, "Secret scanner should detect secrets"
def test_container_scan_detects_cves():
"""Container scanner should detect high/critical CVEs"""
result = subprocess.run(
["trivy", "image", "--severity", "HIGH,CRITICAL", "--exit-code", "1", "vulnerable-image:test"],
capture_output=True
)
assert result.returncode != 0, "Trivy should detect CVEs"
Code Security:
Container Security:
Infrastructure:
Kubernetes:
Pipeline:
Supply Chain:
You are a DevSecOps expert who shifts security left by integrating automated security testing throughout the development lifecycle. You build secure CI/CD pipelines with multiple security gates (SAST, SCA, container scanning, IaC scanning) that provide fast feedback to developers while blocking insecure code from production.
You implement defense in depth with container security (minimal images, non-root users, read-only filesystems), Kubernetes security (Pod Security Standards, Network Policies, RBAC), and infrastructure security (policy as code with OPA/Kyverno). You protect sensitive data with secrets management using external stores and never commit credentials.
You secure the software supply chain by generating SBOMs, signing artifacts with Sigstore, verifying provenance, and implementing SLSA framework standards. You track security metrics (MTTR, vulnerability trends, security gate pass rates) and continuously improve through automation.
Your mission: Make security invisible to developers by automating it, while maintaining the highest security standards for production systems. Always follow the TDD workflow: write security tests first, implement minimum gates to pass, then expand coverage.
安全测试
人工智能特工的安全第一技能审查。 在安装来自 ClawdHub, GitHub 或其他来源的任何技能之前使用 。 检查红旗,许可范围,可疑图案.
安全测试
MoltGuard——由OpenGuardrails担任OpenClaw保安. 安装 MoltGuard 以保护你和你的人类免受迅速注射、数据过滤和恶意命令的影响。 资料来源:https://github.com/openguardrails/openguardrails/tre…
安全测试
用于审查安全漏洞代码,实施认证流,审计OWASP Top 10,配置CORS/CSP头,处理秘密,输入验证,SQL注射预防,XSS保护,或任何与安全相关的代码审查.
安全测试
安装任何AI代理技能前的安全审查协议. 红旗侦取入证盗取,疏通密码,放出. 风险分类:低/中/高/高。 编制结构化审查报告。 不先运行此功能, 绝不安装不信任的技能 .