AGENT SKILL

cloudhsm-cluster-deployer

Provisions AWS CloudHSM clusters with production defaults: cluster creation in a VPC with subnets across 2+ AZs for HA, HSM instance creation, one-time cluster activation (CSR signing, self-signed cert), crypto officer/user management (mu, password policies), HSM backups (daily automatic + on-demand), cross-region backup copy, HA across AZs, PKCS#11/JCE/PCSC library integration, SSL/TLS offload, FIPS 140-2 Level 3 compliance, network security group (sg-xxx), CN-to-cluster-ID mapping, and degradation recovery (create replacement, sync). Emits a READY_TO_DEPLOY checklist with verification commands. Use when creating a CloudHSM cluster, activating an HSM, managing crypto officers, configuring backups, or recovering from HSM degradation. Triggers: create cloudhsm cluster, activate cloudhsm, HSM instance, crypto officer, cloudhsm backup, cross-region backup, PKCS#11, JCE, FIPS 140-2 Level 3, CN-to-cluster-ID.

License 许可
Apache-2.0
最近更新
2026-09-24
Stars
★ 4
Forks
0

01 前言 Front Matter

name
cloudhsm-cluster-deployer
license
Apache-2.0
metadata
{ "tags": "aws, cloudhsm, hsm, cloudops, deploy, security, provisioning, fips-140-2, pkcs11, jce, cluster-activation, hsm-backup", "phase": "1", "author": "Jacky Chan — AWS Community Builder", "domain": "aws-cloudops", "family": "Security", "version": "0.1.0", "keywords": "aws, cloudhsm, hsm, cloudops, deploy, provisioning, security, fips 140-2, pkcs#11, jce, crypto officer, cluster activation, hsm backup, cross-region backup, ha", "task_type": "deploy", "complexity": "high", "entry_point": "false", "skill_class": "capability", "when_to_use": "Invoke when the user wants to create an AWS CloudHSM cluster, activate an HSM (CSR → self-signed cert), create HSM instances across 2+ AZs for HA, manage crypto officers and users, configure backups (daily + on-demand + cross-region copy), integrate PKCS#11/JCE/PCSC libraries, offload SSL/TLS, or recover from HSM degradation. Do NOT invoke for AWS KMS (use kms-key-deployer), ACM (use acm-certificate-deployer), or AWS Signer (use signer-signing-profile-deployer).", "dependencies": "aws-orchestrator", "requires_llm": "true", "verdict_shape": "READY_TO_DEPLOY | PREREQUISITES_MISSING", "lifecycle_status": "active", "supports_pipeline": "true" }
description
Provisions AWS CloudHSM clusters with production defaults: cluster creation in a VPC with subnets across 2+ AZs for HA, HSM instance creation, one-time cluster activation (CSR signing, self-signed cert), crypto officer/user management (mu, password policies), HSM backups (daily automatic + on-demand), cross-region backup copy, HA across AZs, PKCS#11/JCE/PCSC library integration, SSL/TLS offload, FIPS 140-2 Level 3 compliance, network security group (sg-xxx), CN-to-cluster-ID mapping, and degradation recovery (create replacement, sync). Emits a READY_TO_DEPLOY checklist with verification commands. Use when creating a CloudHSM cluster, activating an HSM, managing crypto officers, configuring backups, or recovering from HSM degradation. Triggers: create cloudhsm cluster, activate cloudhsm, HSM instance, crypto officer, cloudhsm backup, cross-region backup, PKCS#11, JCE, FIPS 140-2 Level 3, CN-to-cluster-ID.
compatibility
Agent runtime that reads SKILL.md (Claude Code, Cursor, Windsurf, Codex, Gemini). For live deployment: AWS CLI v2 with cloudhsm, ec2, iam, sts access, plus the CloudHSM client (config -a) and key_mgmt_util / cloudhsm_mgmt_util for crypto officer workflows. Works with Terraform aws_cloudhsm_v2_cluster / aws_cloudhsm_v2_hsm resources and CloudFormation AWS::CloudHSM::Cluster / AWS::CloudHSM::HSM templates.

02 资源文件 19

03 SKILL.md 正文

CloudHSM Cluster Deployer

An AWS CloudOps agent skill that provisions AWS CloudHSM clusters with correct production defaults: cluster creation in a VPC with subnets across 2+ AZs, HSM instance creation, one-time cluster activation (CSR → self-signed cert), crypto officer/user management, backups, HA, PKCS#11/JCE integration, FIPS 140-2 Level 3 posture, and degradation recovery. Captures every configuration decision and emits a READY_TO_DEPLOY checklist with verification commands.

Activation keywords

create CloudHSM cluster, activate CloudHSM, HSM instance, crypto officer, CloudHSM backup, cross-region backup, PKCS#11, JCE, FIPS 140-2 Level 3, CN-to-cluster-ID, HSM degradation recovery.

STRICT output contract

When this skill is invoked with a CloudHSM-provisioning request (create a cluster, activate an HSM, add an HSM instance, configure backups, or recover from degradation), the agent MUST respond with the READY_TO_DEPLOY checklist defined in "Output format" using the literal all-caps labels CLOUDHSM:, VERDICT:, CHECKLIST:, and VERIFICATION_COMMANDS:. Do NOT preface the checklist with prose, headings, or disclaimers — emit the block as the first lines of the response. This contract is what assertion-based evals and downstream provisioning pipelines rely on; deviating from the literal labels breaks automation silently.

If any prerequisite is missing, the verdict is PREREQUISITES_MISSING with a specific gap citation in the checklist (marked [✗]), and READY_TO_DEPLOY MUST NOT also appear. The response uses these literal labels: CLOUDHSM: <cluster-id> (<subnet-az-list>), then VERDICT:, then CHECKLIST: with status markers, then VERIFICATION_COMMANDS: with indented aws cloudhsmv2 ... and aws ec2 ... commands.

Quick navigation

Section When to read
Prerequisites Always — verify before provisioning
Step 1 — Cluster creation (VPC, 2+ AZ subnets) Core topology
Step 2 — HSM instance creation Per-AZ HA
Step 3 — Cluster activation (one-time CSR) One-time gate
Step 4 — Crypto officer/user management CO password + users
Step 5 — HSM backups (daily + on-demand) Backup posture
Step 6 — Cross-region backup copy DR posture
Step 7 — HA across AZs Availability
Step 8 — PKCS#11/JCE/PCSC integration Client libraries
Step 9 — SSL/TLS offload Workload integration
Step 10 — FIPS 140-2 Level 3 compliance Compliance
Step 11 — Network security group (sg-xxx) Network isolation
Step 12 — CN-to-cluster-ID mapping Client discovery
Step 13 — Degradation recovery Replace + sync
NEVER do these things Review before signing off
Output format The literal checklist template
references/cluster-activation-and-users.md Activation + CO detail
references/backups-and-ha.md Backups + HA + recovery detail
references/advanced-patterns.md Heuristics + integration detail
references/error-handling.md API errors + remedies

Mindset

One-line takeaway: An AWS CloudHSM cluster is a single-tenant HSM fleet inside your VPC, validated to FIPS 140-2 Level 3. Creation gives you a cluster-id; creating the FIRST HSM produces a CSR that you sign once to issue the cluster's customer certificate — this activation is one-time and the resulting trust root is what every subsequent HSM inherits. Production clusters need HSMs in ≥2 AZs for HA, and the crypto officer (CO) password is NOT recoverable — lose it (or lose quorum) and the cluster's key material is unrecoverable.

Three misconceptions dominate CloudHSM misdesign at provisioning time:

  • "Create a cluster and you're done." You're not. After create-cluster you must (a) create at least one HSM instance (create-hsm), (b) wait for the HSM to be ACTIVE, (c) sign the CSR the HSM produces and upload the signed cert + your customerCA certificate to activate the cluster. Until activated, no CO, no users, no crypto operations.

  • "One HSM is enough for HA." It isn't. A single HSM is a single point of failure. AWS CloudHSM supports HA by adding HSM instances in different AZs to the SAME cluster — they auto-sync key material. Production needs ≥2 AZs.

  • "The CO password can be reset like a normal password." It cannot. CloudHSM crypto officer credentials are NOT recoverable. Lose the CO password (or the CO user) and you lose access to the key material. Maintain quorum and a secure offline record. No "forgot password" flow exists.

Configuration dependency graph (novel heuristic)

CloudHSM configurations are NOT independent. The cluster needs a VPC with subnets in 2+ AZs. The first HSM instance produces the CSR. Activation (CSR → self-signed cert) is a one-time gate that unlocks CO/user management. Backups are taken per-cluster, copied cross-region for DR. HA requires ≥2 AZs. Use this graph to sequence provisioning.

Configuration Hard dependencies (API error without) Silent failure / immutability Enables downstream
Cluster VPC + ≥2 subnet-AZs; cloudhsm:CreateCluster cluster-id is fixed; subnet list cannot be expanded later cluster-id for HSM creation
HSM instance cluster-id; AZ in subnet list; cloudhsm:CreateHsm first HSM emits a CSR; HSM unreachable until activation CSR for activation
Cluster activation first HSM ACTIVE; CSR signed by in-house CA ONE-TIME — customer cert becomes trust root for every subsequent HSM CO/user mgmt; all crypto ops
Crypto officer (CO) cluster activated; CO logs in via cloudhsm_mgmt_util (mu) CO password NOT recoverable; lose CO + quorum = lose key material user creation, key mgmt, policy
HSM user (CU) cluster activated; CO creates users via mu or key_mgmt_util users are cluster-scoped; same username/password on every HSM (sync) key use via PKCS#11/JCE clients
Daily automatic backup cluster activated; 5-min delta + daily snapshot backups cluster-scoped; retention configurable; hold key material point-in-time restore
On-demand backup cluster activated; via mu createBackup snapshot at time T pre-change safety net
Cross-region backup copy destination region enabled; cloudhsm:CopyBackupToRegion copy is one-way; restores in destination create a new cluster DR in second region
HA across AZs cluster activated; ≥2 HSM instances in different AZs key material auto-syncs across HSMs; client lib load-balances AZ failure tolerance
PKCS#11/JCE client cluster activated; client configured with CN-to-cluster-ID client uses customer CA cert to trust cluster; wrong mapping = failure workload crypto operations
Network security group SG attached to CloudHSM ENIs; inbound from workload subnets SG enforced at ENI; too-broad CIDR = attack surface network isolation
CN-to-cluster-ID mapping registered via configure -a <ip> or DNS CNAME client uses CN to verify cluster identity; mismatch = TLS failure client bootstrap

The activation-is-one-time row is the one a baseline model misses. A naive answer treats activation as repeatable or skips it. In reality the customer certificate you upload after signing the CSR becomes the cluster's permanent trust root; every subsequent HSM inherits it. The CO-password-not-recoverable row is the second most missed: operators assume an AWS-side reset path that does not exist.

Cross-dependency gotchas:

  • Activation is one-time. After uploading the signed customer certificate, you CANNOT re-activate under a different CA without a new cluster (restoring from backup).
  • HA requires ≥2 AZs in the SAME cluster. Two clusters in two AZs do NOT sync key material.
  • Cross-region backup copy creates a restorable backup in the destination region; it does NOT create a running cluster. Restore = restore-backup → new cluster-id in destination.
  • CloudHSM ENI security group must allow inbound from workload subnets on HSM service ports (TCP 22 for SSH-less mu, TCP 17578 for management, TCP 2223 for client). Use narrowest possible CIDR.

Prerequisites (verify before provisioning)

Before provisioning, verify these prerequisites. If any are missing, the verdict is PREREQUISITES_MISSING and the specific gap must be cited.

Prerequisite Why it matters How to verify
VPC + ≥2 subnets in different AZs HA requires ≥2 AZs in the same cluster aws ec2 describe-subnets --filters Name=vpc-id,Values=<vpc>
Security group for CloudHSM ENIs Network isolation; workload subnets only aws ec2 describe-security-groups
cloudhsm:CreateCluster, cloudhsm:CreateHsm IAM Required to create cluster + HSM aws iam simulate-principal-policy ...
Service-linked role (AWSServiceRoleForCloudHSM) CloudHSM auto-creates; if deleted, recreate aws iam get-role --role-name AWSServiceRoleForCloudHSM
HSM type chosen (hsm1.medium) Instance type for the HSM Confirm capacity needs
Backup retention policy decision Daily snapshots; retention in days Confirm policy intent
Cross-region destination (if DR) Destination region must be enabled aws cloudhsmv2 describe-regions-settings --region <dst>
Customer CA (for activation) Bootstrap CA to sign the first CSR Confirm in-house CA or self-sign
CO password storage plan CO password is not recoverable Confirm secret store (Secrets Manager / KMS vault)
CloudHSM client installed (for activation) CloudHSM client + key_mgmt_util + cloudhsm_mgmt_util cloudhsm-cli version

If any prerequisite is missing, output VERDICT: PREREQUISITES_MISSING and cite the specific gap.

Step 1 — Cluster creation (VPC, 2+ AZ subnets)

The cluster is a regional resource that holds HSM instances. Specify the subnet list (one per AZ) and a security group.

# Create the cluster (subnet-az list fixes the HA topology)
CLUSTER_ID=$(aws cloudhsmv2 create-cluster \
  --hsm-type hsm1.medium \
  --subnet-ids subnet-aaa11122 subnet-bbb22233 subnet-ccc33344 \
  --security-group-id sg-cloudhsm-prod \
  --query 'Cluster.ClusterId' --output text --region us-east-1)

# Verify cluster state (CREATE_IN_PROGRESS → UNINITIALIZED)
aws cloudhsmv2 describe-clusters \
  --filters clusterIds=$CLUSTER_ID \
  --query 'Clusters[*].{Id:ClusterId, State:State, Subnets:SubnetMapping, Type:HsmType}' \
  --output table --region us-east-1

Constraint: the subnet list CANNOT be expanded after cluster creation. Plan the AZ topology up front. Adding a new AZ requires a new cluster.

Step 2 — HSM instance creation

Create at least one HSM (per AZ) in the cluster. The first HSM emits the CSR for activation.

# Create HSMs — one per AZ for HA
aws cloudhsmv2 create-hsm --cluster-id "$CLUSTER_ID" --availability-zone us-east-1a \
  --ip-address 10.0.1.10 --region us-east-1
aws cloudhsmv2 create-hsm --cluster-id "$CLUSTER_ID" --availability-zone us-east-1b \
  --ip-address 10.0.2.10 --region us-east-1

# Verify each HSM reaches ACTIVE
aws cloudhsmv2 describe-clusters --filters clusterIds=$CLUSTER_ID \
  --query 'Clusters[0].Hsms[*].{Id:HsmId, AZ:AvailabilityZone, State:State, ENI:EniIp}' \
  --output table --region us-east-1

Each HSM gets an ENI in the specified subnet. State transitions: CREATE_IN_PROGRESS → ACTIVE. Once the first HSM is ACTIVE, the CSR is available via describe-clusters.

Step 3 — Cluster activation (one-time CSR → self-signed cert)

Activation is the one-time gate that unlocks CO/user management.

# 1. Fetch the CSR from the first ACTIVE HSM
aws cloudhsmv2 describe-clusters \
  --filters clusterIds=$CLUSTER_ID \
  --query 'Clusters[0].Certificates.ClusterCertificate' \
  --output text --region us-east-1 > cluster.csr

# 2. Sign the CSR with your in-house CA (or self-sign for bootstrap)
openssl x509 -req -days 3650 -in cluster.csr \
  -CA customerCA.crt -CAkey customerCA.key -set_serial 01 \
  -out customerCertificate.crt

# 3. Upload signed cert + CA cert to activate the cluster
aws cloudhsmv2 initialize-cluster \
  --cluster-id "$CLUSTER_ID" \
  --signed-cert fileb://customerCertificate.crt \
  --trust-anchor fileb://customerCA.crt --region us-east-1

# 4. Verify state transitions to INITIALIZED → ACTIVE
aws cloudhsmv2 describe-clusters \
  --filters clusterIds=$CLUSTER_ID \
  --query 'Clusters[0].{State:State, Cert:Certificates}' \
  --output table --region us-east-1

Critical: activation is ONE-TIME. After this, the customer certificate is the trust root. To change CAs, restore from backup into a new cluster.

Step 4 — Crypto officer/user management

Full procedure: cluster-activation-and-users.md. Log in via cloudhsm_mgmt_util (mu) over the ENI; create the first CO (admin) and CUs. The CO password is the crown-jewel secret — store in Secrets Manager (customer-managed KMS key) and keep >=2 COs for quorum.

Step 5 — HSM backups (daily + on-demand)

Full procedure: backups-and-ha.md. Automatic backups every ~5 minutes (delta) plus a daily snapshot; on-demand backups via the mu createBackup command. Backups are cluster-scoped; retention is configurable per cluster.

Step 6 — Cross-region backup copy

Full procedure: backups-and-ha.md. copy-backup-to-region for DR — the copy is one-way and restore in the destination creates a NEW cluster-id (cold DR, not a hot replica).

Step 7 — HA across AZs

Full procedure: backups-and-ha.md. 2 AZs tolerates 1 AZ failure (minimum production posture); 3 AZs tolerates 1 failure during a replace (recommended). Two separate clusters do NOT sync key material.

Step 8 — PKCS#11/JCE/PCSC library integration

Full procedure: advanced-patterns.md. The client uses the customer CA certificate (uploaded at activation) to trust the cluster — keep the bundle current on every client host.

Step 9 — SSL/TLS offload

Full procedure: advanced-patterns.md. Generate the keypair inside the HSM; the web server holds only the certificate — the private key never leaves the HSM.

Step 10 — FIPS 140-2 Level 3 compliance

Full procedure: advanced-patterns.md. In-HSM operations only; CloudTrail control-plane + HSM data-plane audit trail; narrow SG inbound; CO quorum for destructive operations.

Step 11 — Network security group (sg-xxx)

Full procedure: advanced-patterns.md. Inbound TCP 2223 (client) from workload subnets only; TCP 17578 (management utility); TCP 22 (mu). Use the narrowest CIDR — never 0.0.0.0/0.

Step 12 — CN-to-cluster-ID mapping

Full procedure: cluster-activation-and-users.md. Register the mapping via configure -a <eni-ip> or a DNS CNAME; a wrong mapping fails the client TLS handshake.

Step 13 — Degradation recovery (create replacement, sync)

Full procedure: backups-and-ha.md. Create the replacement HSM FIRST, wait for ACTIVE + sync, then delete the degraded HSM — never drop below 1 ACTIVE HSM.

NEVER do these things

  1. NEVER deploy a production CloudHSM cluster with HSMs in only one AZ. Single-AZ is a single point of failure. Use ≥2 AZs.

  2. NEVER skip cluster activation. create-cluster + create-hsm leaves the cluster UNINITIALIZED. Without activation (CSR → signed cert upload), no CO, no users, no crypto.

  3. NEVER treat activation as repeatable. The customer certificate is the cluster's permanent trust root. To change CAs, restore from backup into a new cluster.

  4. NEVER lose the CO password. CloudHSM has NO password reset flow. Lose CO + lose quorum = lose key material. Maintain ≥2 COs and a secure offline record.

  5. NEVER use two clusters as HA. Two clusters do NOT sync key material. HA is intra-cluster: ≥2 HSMs in different AZs in the SAME cluster.

  6. NEVER expand the subnet-AZ list after cluster creation. The subnet list is fixed at create-cluster time. Adding an AZ requires a new cluster.

  7. NEVER open CloudHSM ENI security group to 0.0.0.0/0. Use the narrowest workload-subnet CIDR.

  8. NEVER export key material to workload memory if you need FIPS 140-2 Level 3. Use in-HSM operations via PKCS#11/JCE.

  9. NEVER assume cross-region backup copy creates a hot replica. Copy is one-way. Restore creates a NEW cluster-id; cold DR, not a running failover target.

  10. NEVER delete the degraded HSM before the replacement is ACTIVE. Dropping below 1 ACTIVE HSM loses HA. Create replacement first; verify sync; then delete.

  11. NEVER confuse CloudHSM with KMS. CloudHSM is single-tenant hardware, FIPS 140-2 Level 3, BYO client libraries, you manage users/keys. KMS is multi-tenant regional, FIPS 140-2 Level 2 (mostly), AWS manages the HSM backing.

Output format

CLOUDHSM: <cluster-id> (<az-list>)
VERDICT: READY_TO_DEPLOY | PREREQUISITES_MISSING
CHECKLIST:
  [✓|✗] Cluster ID: <cluster-id>
  [✓|✗] VPC + subnets: <vpc-id> (subnet-AZ list)
  [✓|✗] HSM type: hsm1.medium
  [✓|✗] HSM instances: <count> across <n> AZs (≥2 for HA)
  [✓|✗] Cluster state: UNINITIALIZED → ACTIVE (after activation)
  [✓|✗] Activation: CSR signed by <CA>, customer cert uploaded (ONE-TIME)
  [✓|✗] Customer CA: <ca-name> (trust root for the cluster)
  [✓|✗] Crypto officer (CO): ≥2 CO users created
  [✓|✗] CO password: stored in Secrets Manager (KMS-encrypted)
  [✓|✗] Crypto users (CU): <cu-list>
  [✓|✗] Password policy: <policy-summary>
  [✓|✗] Daily automatic backup: enabled (cluster-scoped)
  [✓|✗] On-demand backup: pre-change snapshot available
  [✓|✗] Cross-region backup copy: destination region <region>
  [✓|✗] HA: <n> AZs (tolerates 1 AZ failure)
  [✓|✗] PKCS#11/JCE client: configured with CN-to-cluster-ID mapping
  [✓|✗] SSL/TLS offload: <workload-list>
  [✓|✗] FIPS 140-2 Level 3: in-HSM operations, audit trail configured
  [✓|✗] Security group: sg-cloudhsm-prod (narrow CIDR inbound)
  [✓|✗] CN-to-cluster-ID mapping: configured (DNS CNAME or configure -a)
  [✓|✗] Degradation recovery: documented (create replacement, sync, delete)
  [✓|✗] Tags: <key=value list>
VERIFICATION_COMMANDS:
  aws cloudhsmv2 describe-clusters --filters clusterIds=<cluster-id> --region <region>
  aws cloudhsmv2 describe-backups --filters clusterIds=<cluster-id> --region <region>
  aws ec2 describe-security-groups --group-ids <sg-id> --region <region>
  aws cloudtrail lookup-events --lookup-attributes AttributeKey=EventSource,AttributeValue=cloudhsm.amazonaws.com --region <region>

Worked example — 3-AZ CloudHSM cluster with activation and HA

CLOUDHSM: cluster-abc123def (us-east-1a, us-east-1b, us-east-1c)
VERDICT: READY_TO_DEPLOY
CHECKLIST:
  [✓] Cluster ID: cluster-abc123def
  [✓] VPC + subnets: vpc-aaa11122 (subnet-aaa in 1a, subnet-bbb in 1b, subnet-ccc in 1c)
  [✓] HSM type: hsm1.medium
  [✓] HSM instances: 3 across 3 AZs (tolerates 1 AZ failure during replace)
  [✓] Cluster state: ACTIVE
  [✓] Activation: CSR signed by in-house customerCA, cert uploaded (ONE-TIME)
  [✓] Customer CA: customerCA (trust root for cluster)
  [✓] Crypto officer (CO): admin, officer2 (2 COs for quorum)
  [✓] CO password: stored in Secrets Manager (KMS-encrypted)
  [✓] Crypto users (CU): app-tls-user, app-code-sign-user
  [✓] Password policy: min 14 chars, mixed, 90-day rotation
  [✓] Daily automatic backup: enabled (5-min delta + daily snapshot)
  [✓] On-demand backup: pre-change snapshot backup-aaa11122
  [✓] Cross-region backup copy: us-west-2
  [✓] HA: 3 AZs (tolerates 1 AZ failure during replace)
  [✓] PKCS#11 client: configured (CN-to-cluster-ID via DNS CNAME)
  [✓] SSL/TLS offload: nginx using CloudHSM-held private key
  [✓] FIPS 140-2 Level 3: in-HSM ops, CloudTrail + HSM logs
  [✓] Security group: sg-cloudhsm-prod (10.0.10.0/24 inbound on 2223)
  [✓] CN-to-cluster-ID mapping: cloudhsm.internal.example.com → 10.0.1.10
  [✓] Degradation recovery: documented (create replacement, sync, delete)
  [✓] Tags: Environment=production, Workload=tls-offload, Owner=platform-team
VERIFICATION_COMMANDS:
  aws cloudhsmv2 describe-clusters --filters clusterIds=cluster-abc123def --region us-east-1
  aws cloudhsmv2 describe-backups --filters clusterIds=cluster-abc123def --region us-east-1
  aws ec2 describe-security-groups --group-ids sg-cloudhsm-prod --region us-east-1
  aws cloudtrail lookup-events --lookup-attributes AttributeKey=EventSource,AttributeValue=cloudhsm.amazonaws.com --region us-east-1

References (load on demand)

  • Cluster activation and users — one-time CSR activation flow, CO/CU management via mu, password policies, quorum, CN-to-cluster-ID mapping (Steps 4 and 12 in detail)
  • Backups and HA — backup model, cross-region copy, HA topology, degradation recovery, DR drill (Steps 5, 6, 7, and 13 in detail)
  • Advanced patterns — expert heuristics (one-time activation, >=2 AZs, CO password non-recoverability), PKCS#11/JCE/PCSC integration, SSL/TLS offload, FIPS 140-2 Level 3 posture, security groups (Steps 8-11 in detail)
  • Error handling — API error codes and remedies
  • End-to-end walkthrough — full worked example from invocation to verification

Domain

AWS CloudOps / AWS CloudHSM Cluster Provisioning & HSM Lifecycle.

AWS documentation

04 同仓库更多 Skills

jhkchan
jhkchan/aws-agent-skills

xray-tracing-deployer

Deploys AWS X-Ray distributed tracing with production-grade configuration: X-Ray daemon deployment (EC2 systemd, ECS sidecar, EKS DaemonSet, Lambda built-in), SDK instrumentation (Java, Python, Node.js, Go, .NET) with annotations and metadata, sampling rules (default reservoir + rate vs custom rules with URL / method / service-name predicates), service map generation, trace query and retrieval, groups and insights, CloudWatch ServiceLens integration, IAM permissions (PutTraceSegments, GetSamplingRules), and latest features (Lambda Powertools tracing, OpenTelemetry / ADOT Collector support, W3C trace context). Emits a READY_TO_DEPLOY checklist with every configuration item verified. Use when enabling X-Ray on a new application, adding tracing to ECS / EKS / Lambda, validating a sampling configuration, or generating daemon deployment CLI and IaC templates. Triggers: X-Ray, distributed tracing, X-Ray daemon, sampling rules, service map, X-Ray SDK, OpenTelemetry, ADOT, Lambda Powertools, CloudWatch ServiceLens.

★ 4
jhkchan
jhkchan/aws-agent-skills

wellarchitected-workload-auditor

Audits AWS Well-Architected Tool workloads for review staleness, per-pillar high-risk issue counts, milestone tracking gaps, and remediation plan completeness. Emits a deterministic verdict (STALE_REVIEW | HIGH_RISK | CONFIG_GAP | OK) per workload with per-pillar risk breakdown, milestone trend analysis, and specific remediation. Use when reviewing Well-Architected workloads, checking WA review freshness, auditing high-risk issues per pillar, validating milestone coverage, or tracking remediation plans.

★ 4
jhkchan
jhkchan/aws-agent-skills

wellarchitected-review-operator

Operates AWS Well-Architected Tool reviews end-to-end — creates workloads, runs pillar reviews (operational excellence, security, reliability, performance, cost optimization, sustainability, Prosperity), answers questions with risk-tier rationale, generates reports and improvement plans, creates milestones, and integrates with Trusted Advisor and Well-Architected Labs. Runs deterministic pre-checks (workload existence, lens availability, pillar coverage), emits the exact update-answer / get-consolidated-report CLI behind a CONFIRM gate, and verifies risk-tier transitions. Emits READY | BLOCKED | COMPLETED. Use when creating a workload, answering pillar questions, generating an improvement plan, creating a milestone, or integrating with Trusted Advisor. Triggers: well-architected tool, well architected review, workload review, pillar, operational excellence, security, reliability, performance, cost optimization, sustainability, prosperity, improvement plan, milestone, trusted advisor.

★ 4
jhkchan
jhkchan/aws-agent-skills

wafv2-web-acl-deployer

Provisions AWS WAFv2 Web ACLs with correct production defaults: scope selection (CLOUDFRONT vs REGIONAL), managed rule groups (AWSManagedRulesCommonRuleSet, SQLiRuleSet, LinuxRuleSet, WindowsRuleSet, IPReputationList, BotControlRuleSet, AmazonIpReputationList, ATP), custom rules (byte match, IP set, regex pattern, geo match, size constraint, rate-based), rule priority ordering, logging destinations (Kinesis Firehose, CloudWatch Logs, S3), resource association (ALB, API Gateway, CloudFront), rate-based rules with aggregate keys (IP, forwarded IP), Challenge and CAPTCHA actions, and account takeover prevention. Emits a READY_TO_DEPLOY checklist. Use when creating a Web ACL, attaching managed rules, configuring rate limiting, enabling WAF logging, associating a Web ACL with an ALB/API Gateway/CloudFront, or configuring CAPTCHA/Challenge actions. Triggers: create WAF Web ACL, WAFv2, managed rule groups, WAF rate limiting, WAF logging, WAF CAPTCHA, WAF Challenge action, bot control, ATP managed rules.

★ 4
jhkchan
jhkchan/aws-agent-skills

wafv2-web-acl-auditor

Audits AWS WAFv2 Web ACL configurations to determine whether the ACL provides effective protection: default-action posture (Allow vs Block), managed-rule-group coverage gaps, rule effectiveness (BLOCK vs COUNT, shadow/bypass rules, stale exclusions), rate-based rule correctness, logging and visibility configuration, and text-transformation bypass vectors. Emits a deterministic verdict per Web ACL. Use when reviewing a WAFv2 Web ACL for security posture, checking managed-rule coverage, validating rate-limit configuration, auditing WAF logging, or hardening a WAF before production deployment. Triggers: WAF, WAFv2, Web ACL, managed rule group, AWSManagedRules, rate-based rule, CAPTCHA, Challenge, OverrideAction, Count mode, forwarded IP, Bot Control, ATP, text transformation, WAF logging, WAF visibility, OWASP, SQLi, XSS, RFI.

★ 4
jhkchan
jhkchan/aws-agent-skills

waf-rule-deployer

Provisions AWS WAFv2 rule sets with production defaults: Web ACL creation (CloudFront scope in us-east-1 vs Regional scope), rule groups (managed vs custom) with capacity planning in WCU (Web ACL Capacity Units, 1500 default per ACL), managed rule groups (AWSManagedRulesCommonRuleSet, AWSManagedRulesSQLiRuleSet, AWSManagedRulesAmazonIpReputationList, AWSManagedRulesLinuxRuleSet, AWSManagedRulesWindowsRuleSet, AWSManagedRulesUnixRuleSet), custom rules (byte-match, regex pattern set, geo-match, rate-based), IP set and regex pattern set resources, rule priority ordering (lowest number evaluated first), action types (allow, block, count, CAPTCHA, Challenge), bot control managed rule group, account takeover. Triggers: create WAF rule, WAF managed rule group, WAF custom rule, WAF byte match, WAF geo match, WAF rate based rule, WAF IP set, WAF regex pattern set, WAF WCU budget, WAF CloudFront ACL, WAF logging Kinesis Firehose, WAF bot control, WAF ATP, WAF label matching, WAF Challenge action, WAF CAPTCHA.

★ 4