
Cloud Security Engineer
- 28 installs
- 7 repo stars
- Updated May 20, 2026
- daemon-blockint-tech/agentic-enteprises-skill
Implement and audit cloud security: org guardrails, IAM hardening, network segmentation, KMS encryption, CSPM detective controls, and misconfig fixes.
About
Guides cloud security engineering on AWS/GCP/Azure covering org guardrails, IAM hardening, network segmentation, encryption/KMS, audit logging, CSPM, and misconfiguration remediation. A developer uses it when implementing or auditing cloud security controls.
- Org guardrails (SCPs/org policies), cloud IAM hardening, KMS encryption
- CSPM and detective controls: Config, Security Hub, GuardDuty, SCC, Defender
Cloud Security Engineer by the numbers
- 28 all-time installs (skills.sh)
- Ranked #1,512 of 2,203 Security skills by installs in the Skillselion catalog
- Data as of Jul 29, 2026 (Skillselion catalog sync)
npx skills add https://github.com/daemon-blockint-tech/agentic-enteprises-skill --skill cloud-security-engineerAdd your badge
Show developers this skill is listed on Skillselion. Paste this into your README.
| Installs | 28 |
|---|---|
| repo stars | ★ 7 |
| Last updated | May 20, 2026 |
| Repository | daemon-blockint-tech/agentic-enteprises-skill ↗ |
What it does
Implement and audit cloud security: org guardrails, IAM hardening, network segmentation, KMS encryption, CSPM detective controls, and misconfig fixes.
Files
Cloud Security Engineer
When to Use
- Design and implement org/account guardrails — SCPs, policy constraints, landing zone security
- Harden cloud IAM — roles, trust policies, permission boundaries, federation, break-glass
- Secure cloud networking — segmentation, SG/NSG rules, private endpoints, egress control
- Configure encryption — KMS/CMK policies, default encryption, TLS, secrets managers
- Enable audit and detective controls — CloudTrail/Audit Logs, Config, GuardDuty, CSPM
- Remediate misconfigurations from scans, audits, or Well-Architected security pillar
- Review workload designs for cloud threat patterns (IMDS, public buckets, open SGs)
- Integrate cloud findings into vulnerability and exception workflows
- Support incident forensics with cloud log analysis (with SOC/IR partners)
When NOT to Use
- Company security strategy, policies, board metrics →
cybersecurity - SSO/PAM/SIEM/EDR for corp-wide stack (non-cloud-specific) →
information-security-engineer - SAST/SCA/SBOM and GitHub Actions hardening →
devsecops - Live SOC alert triage and playbooks →
soc-analyst - Cloud telemetry threat hunts and ATT&CK campaigns →
threat-hunter - Authorized exploitation and pentest validation →
penetration-tester - Network/AD/infra pentest from corp paths →
network-pentester - Web/API OWASP testing →
web-pentester - GRC program, audit prep, vendor questionnaires →
compliance-specialist - SOC 2 control narratives and audit binders →
compliance-engineer,cloud-compliance-specialist - Build VPC/RDS without security as primary goal →
cloud-engineer - Landing zone business architecture and migration →
cloud-architect - Cloud program strategy and CCoE investment themes →
vp-of-cloud - Product multi-tenant isolation in app layer →
product-infrastructure-security-engineer - Cloud access tickets and patching →
cloud-system-administrator - Entitlement design, access reviews, federation, PAM →
iam-specialist - Customer security questionnaires, deal compliance fit (architecture) →
solutions-architect
Related skills
| Need | Skill |
|---|---|
| VP cloud program and risk investment themes | vp-of-cloud |
| Corporate security tooling and IdP | information-security-engineer |
| Pipeline and supply-chain security | devsecops |
| Cloud architecture and WAF reviews | cloud-architect |
| Enterprise CCoE and regulated program | enterprise-cloud-architect |
| Cloud resource implementation | cloud-engineer |
| Terraform platform modules | infrastructure-engineer |
| GRC program, gap plans, audit coordination | compliance-specialist |
| Compliance evidence (org-wide) | compliance-engineer |
| Cloud audit evidence and framework mapping | cloud-compliance-specialist |
| SOC triage and playbooks | soc-analyst |
| Active security IR, cloud log coordination | incident-responder |
| Cloud telemetry threat hunts and hunt campaigns | threat-hunter |
| Cloud alert investigation and detection tuning | defensive-security-analyst |
| Pentest validation | penetration-tester |
| Network/AD/infra pentest | network-pentester |
| Web/API OWASP pentest | web-pentester |
| Product tenancy | product-infrastructure-security-engineer |
| Customer deal security/compliance fit memo | solutions-architect |
| CVD and disclosure | technical-program-manager-security-cvd |
| Cloud audit log forensics and super-timelines after preservation | digital-forensics-analyst |
| Security risk registers and treatment prioritization | security-risk-analyst |
| IAM lifecycle, access reviews, federation, PAM | iam-specialist |
Core Workflows
1. Scope and shared responsibility
Cloud security boundaries, provider vs customer duties.
See `references/cloud_security_scope.md`.
2. Cloud IAM and identity
Roles, federation, privilege escalation prevention.
See `references/identity_iam_cloud.md`.
3. Network security in cloud
Segmentation, private access, logging.
See `references/network_cloud_security.md`.
4. Data protection and KMS
Encryption, keys, secrets.
See `references/data_encryption_kms.md`.
5. Logging, CSPM, and detection
Audit logs, posture management, native detectors.
See `references/detection_cspm_logging.md`.
6. Architecture review and remediation
Threat patterns, review checklist, fix prioritization.
See `references/secure_cloud_architecture_review.md`.
Outputs
- Guardrail definition — SCP/policy JSON, exceptions, rollout plan
- IAM policy set — least-privilege roles with trust boundaries documented
- Network security diagram — zones, flows allowed/denied, private endpoints
- Remediation backlog — finding, severity, owner, compensating control
- Control evidence — Config rules, scan exports, sample audit log queries
- Architecture review notes — risks, required controls before launch
Principles
- Deny by default — explicit allow for network and IAM
- Security as code — guardrails versioned and reviewed like application code
- Detect and prove — every preventive control has a detective check
- Break-glass is rare and monitored — not a bypass for convenience
- Minimize blast radius — account segmentation and permission boundaries
Cloud security scope
Table of contents
1. Shared responsibility 2. Role boundary 3. Multi-account model 4. Compliance hooks
Shared responsibility
| Layer | Customer typically owns | Provider owns |
|---|---|---|
| Data classification | Yes | — |
| IAM and access | Yes | — |
| Network config in VPC | Yes | Physical network |
| OS and app on IaaS | Yes | Hypervisor |
| Managed service config | Yes | Service software |
| Physical DC | — | Yes |
Document which controls apply per workload (IaaS vs PaaS vs SaaS).
Role boundary
| cloud-security-engineer | Partner skill |
|---|---|
| SCPs, org policies, Config rules | cloud-architect designs landing zone shape |
| Cloud IAM hardening | cloud-system-administrator executes access tickets |
| CSPM remediation | cloud-engineer implements non-security infra |
| Pipeline OIDC, image scan gates | devsecops |
| Corp IdP, EDR, email security | information-security-engineer |
| Audit binder and control mapping | compliance-engineer |
Multi-account model
Typical secure structure:
org root (SCPs)
├── security / audit (log archive, Security Hub admin)
├── shared services (DNS, egress, CI artifacts)
├── workload prod / non-prod (separate accounts)
└── sandbox (strict SCP, auto-cleanup)Principles:
- No workloads in management account
- Centralized logging — immutable log archive account
- Separate security tooling account for delegated admin
- SCPs deny dangerous APIs (disable root keys, restrict regions)
Compliance hooks
Implement technical controls; partner with compliance-engineer for:
- Control ID mapping (SOC 2, ISO 27001, CIS)
- Evidence retention periods
- Exception register with expiry
Use CIS Benchmarks and provider Well-Architected Security pillar as baselines—not as sole authority without business context.
Data protection and KMS
Table of contents
1. Encryption at rest 2. Encryption in transit 3. Key management 4. Secrets
Encryption at rest
Defaults to enforce:
- Block storage — encrypted volumes/disks; no unencrypted snapshots shared
- Object storage — SSE-KMS or CMEK; block public ACLs/buckets at org level
- Databases — TDE or managed encryption; no public endpoints
- Backups — encrypted; cross-account copy with key policy
SCP/org policy: deny s3:PutObject without encryption headers where applicable.
Encryption in transit
- TLS 1.2+ on all customer and internal APIs
- Managed certs (ACM, etc.) with auto-renewal
- mTLS for high-risk service mesh where required
- Disable legacy protocols on load balancers
Key management
| Control | Practice |
|---|---|
| CMK ownership | Security or dedicated keys account |
| Key policy | Least privilege; separate encrypt vs admin |
| Rotation | Annual or on compromise; document process |
| Separation of duties | Key admins ≠ data admins |
| Audit | CloudTrail KMS events to SIEM |
Avoid hardcoded keys in AMIs, user-data, or templates — scan in CI (devsecops).
Secrets
- Secrets Manager / Parameter Store / Key Vault — not plain env in task defs
- Rotation lambdas or native rotation enabled
- RBAC on secret read; audit
GetSecretValue - No secrets in CloudFormation/Terraform state unencrypted — use references + backend encryption
App-layer secret design for multi-tenant products → product-infrastructure-security-engineer.
Logging, CSPM, and detection
Table of contents
1. Audit logging 2. Posture management 3. Native threat detection 4. SIEM integration
Audit logging
Org-wide mandatory:
| Provider | Control |
|---|---|
| AWS | CloudTrail org trail, all regions, log file validation, S3 + KMS |
| GCP | Admin Activity + Data Access (where policy allows), sinks to central bucket |
| Azure | Activity Log to Log Analytics / storage; diagnostic settings on resources |
Protect logs:
- Immutable storage; separate audit account
- Alert on trail disable, policy change, root login
- Retention meets legal and
compliance-engineerrequirements
Posture management
CSPM / native posture tools:
- AWS Security Hub + Config rules (CIS conformance packs)
- GCP Security Command Center
- Azure Defender for Cloud + regulatory compliance dashboard
Workflow:
1. Enable org-wide with delegated admin account 2. Baseline rules — fail on critical misconfigs (public S3, open SG, no MFA on root) 3. Route findings to ticketing; SLA by severity 4. Track mean time to remediate; exception register with expiry
Native threat detection
Enable and tune:
- GuardDuty / Defender for Cloud / SCC threat detectors
- VPC Flow–based anomaly where available
- Integrate with SOAR playbooks (
defensive-security-analystfor alert handling)
Tune suppressions carefully; document false positives.
SIEM integration
Forward to corporate SIEM (information-security-engineer):
- Normalize fields (account, region, principal, resource ARN)
- High-value detections: IAM policy change, SG open to world, KMS disable, snapshot public
- Correlate with IdP sign-in and EDR where available
Cloud-security-engineer owns cloud log pipeline quality; SOC owns detection logic and response tiers.
Cloud IAM and identity
Table of contents
1. Human access 2. Workload identity 3. Privilege escalation paths 4. Reviews and break-glass
Human access
- Federate console/API via corporate IdP (
information-security-engineerfor IdP design) - MFA required for all human principals; no password-only IAM users
- Permission sets (AWS SSO) or group-based RBAC — not direct admin on accounts
- Session duration limits; no permanent access keys for humans
Workload identity
| Pattern | Use |
|---|---|
| Instance/pod role | VMs, Lambda, GKE/EKS service accounts |
| OIDC federation | CI/CD from GitHub/GitLab (devsecops) |
| Cross-account role | Hub-spoke with external ID and tight trust |
Rules:
- One role per workload function; no shared “super” role
- Permission boundaries on roles that can create IAM
- Scope resource ARNs in policies; avoid
*onActionandResourcetogether
Privilege escalation paths
Audit regularly for:
| Pattern | Risk |
|---|---|
iam:PassRole + ec2:RunInstances | Admin instance |
lambda:CreateFunction + pass admin role | Backdoor function |
sts:AssumeRole on * in trust | Lateral movement |
Overly broad kms:Decrypt | Data exfil |
| Public AMIs or snapshots | Data leak |
Use IAM Access Analyzer, policy simulation, and pentest findings (offensive-security-analyst).
Reviews and break-glass
- Quarterly access review with evidence export per account
- Break-glass role: MFA, short session, alert on use, post-use review
- Remove unused roles and inline policies; prefer managed policy versions
Operational grants → cloud-system-administrator; security standard → this skill.
Network security in cloud
Table of contents
1. Segmentation 2. Ingress and egress 3. Private connectivity 4. Logging and inspection
Segmentation
Design tiers:
| Tier | Typical placement |
|---|---|
| Public edge | ALB/Cloud LB only; no admin |
| App | Private subnets; no direct internet |
| Data | Private; SG/NSG allow only app tier |
| Management | Bastion-less preferred; SSM/serial console |
Avoid flat VPCs with wide SG rules (0.0.0.0/0 on SSH/RDP/DB ports).
Ingress and egress
Ingress:
- WAF on public HTTP(S) where applicable
- Default deny SG/NSG; document each allow rule owner
- No admin ports on
0.0.0.0/0
Egress:
- Restrict outbound to required FQDNs/IPs where possible (egress firewall, NAT with filtering)
- Block metadata abuse — IMDSv2 required on AWS; restrict metadata access on containers
Private connectivity
Prefer over public internet for service-to-service:
- AWS PrivateLink / GCP Private Service Connect / Azure Private Link
- VPC/VNet peering or Transit Gateway — document routing and NACL/SG at both ends
- Hybrid: VPN or Direct Connect/Interconnect with encryption
Logging and inspection
Enable and retain:
| Log | Purpose |
|---|---|
| VPC Flow Logs / equivalent | Connection audit |
| DNS query logs | Exfil and C2 patterns |
| LB access logs | HTTP abuse |
| Firewall logs | Allow/deny proof |
Centralize to security log archive; integrate per detection_cspm_logging.md.
Optional: IDS/IPS, NGFW appliances — architecture sign-off with cloud-architect.
Secure cloud architecture review
Table of contents
1. Review triggers 2. Checklist 3. Common misconfigurations 4. Remediation prioritization
Review triggers
Run review before:
- New workload landing in prod account
- Internet-exposed service launch
- Cross-account trust or peering
- Handling regulated or PII data
- Major migration wave (
cloud-architectcoordinates)
Checklist
| Area | Verify |
|---|---|
| Identity | Least privilege; no long-lived keys; federation |
| Network | Segmentation; no public DB; private endpoints |
| Data | Encryption at rest/transit; backup encrypted |
| Logging | Audit on; flows enabled; central retention |
| Resilience | Not a security substitute — align with site-reliability-engineer for availability |
| Supply chain | Signed images; private registry (devsecops) |
| Admin | No console on prod data plane; break-glass only |
Pair with Well-Architected Security pillar and threat model for critical systems.
Common misconfigurations
| Finding | Typical fix |
|---|---|
| Public S3/GCS/Blob | Block public access org-wide; fix ACLs |
| Open security group | Restrict source; use bastion-less admin |
| Overprivileged role | Split roles; permission boundary |
| Unencrypted volume | Enable default encryption; re-encrypt |
| CloudTrail off | Enable org trail; alert on disable |
| IMDSv1 | Require IMDSv2 |
| Wildcard trust policy | Scope account IDs + external ID |
| Keys in git | Rotate; secrets manager; scan in CI |
Validate fixes with Config rule or rescan — not only console spot check.
Remediation prioritization
| Priority | Criteria |
|---|---|
| P0 | Exploitable internet exposure of data or admin |
| P1 | Credential risk, broad IAM, missing audit |
| P2 | Encryption gap on non-public data |
| P3 | Hardening drift, low-risk hygiene |
Document compensating controls for accepted risk: owner, expiry, detection in place.
Pentest validation of fixes → offensive-security-analyst.