AGENT SKILL

rds-backup-restore-operator

Operates RDS and Aurora backup and restore workflows safely — automated backup window configuration, manual snapshots for pre-maintenance safety, cross-region and cross-account snapshot copy, point-in-time restore (PITR), Aurora Backtrack for in-place rewind, S3 export/import, and full post- restore verification. Runs deterministic pre-checks (subnet group, security group, option group, KMS key, instance status), executes the operation behind a CONFIRM gate, and emits a verdict (READY | BLOCKED | COMPLETED) per operation with the exact CLI sequence, expected side-effects (new instance endpoint, connection-string updates), and verification commands. Use when configuring RDS backup retention, creating pre-migration snapshots, restoring to a point in time, backtracking an Aurora cluster, exporting a snapshot to S3, or recovering from a bad change.

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

01 前言 Front Matter

name
rds-backup-restore-operator
license
Apache-2.0
metadata
{ "tags": "rds, aurora, databases, backup, restore, pitr, backtrack, snapshot, recovery, disaster-recovery", "phase": "4", "author": "Jacky Chan — AWS Community Builder", "domain": "aws-cloudops", "family": "Databases", "version": "0.1.0", "keywords": "RDS, Amazon Aurora, Aurora MySQL, Aurora PostgreSQL, automated backup, manual snapshot, point-in-time recovery, PITR, backtrack, fast database clone, cross-region snapshot copy, cross-account snapshot share, S3 export, S3 import, backup retention, backup window, maintenance window, Multi-AZ, restore-db-instance-to-point-in-time, restore-db-instance-from-db-snapshot, backtrack-db-cluster, create-db-snapshot, start-export-task", "task_type": "operate", "complexity": "medium", "entry_point": "false", "skill_class": "capability", "when_to_use": "Configuring RDS backup retention or backup window, creating a manual snapshot before maintenance or a risky change, copying snapshots cross- region or cross-account, restoring an RDS instance to a point in time, restoring from a snapshot, backtracking an Aurora MySQL cluster, exporting a snapshot to S3, importing from S3 into Aurora MySQL, or recovering from a bad schema change.", "requires_llm": "true", "verdict_shape": "READY | BLOCKED | COMPLETED", "invocation_schema": "Input: either (a) an RDS instance/cluster configuration with the intended operation (create-snapshot, pitr-restore, snapshot-restore, backtrack, export, import), OR (b) an instance-id / cluster-id + operation for live-account execution. Output: deterministic OPERATION/VERDICT/PRE_CHECKS/ STEPS/POST_VERIFY block per operation, where VERDICT is one of READY, BLOCKED, COMPLETED.", "supports_pipeline": "true", "activation_triggers": "configure RDS backup retention, create RDS manual snapshot, restore RDS to point in time, PITR restore RDS, restore RDS from snapshot, backtrack Aurora cluster, Aurora fast clone, copy RDS snapshot cross-region, share RDS snapshot cross-account, export RDS snapshot to S3, import S3 into Aurora MySQL, RDS pre-migration snapshot, recover from bad RDS change, verify RDS restore" }
description
Operates RDS and Aurora backup and restore workflows safely — automated backup window configuration, manual snapshots for pre-maintenance safety, cross-region and cross-account snapshot copy, point-in-time restore (PITR), Aurora Backtrack for in-place rewind, S3 export/import, and full post- restore verification. Runs deterministic pre-checks (subnet group, security group, option group, KMS key, instance status), executes the operation behind a CONFIRM gate, and emits a verdict (READY | BLOCKED | COMPLETED) per operation with the exact CLI sequence, expected side-effects (new instance endpoint, connection-string updates), and verification commands. Use when configuring RDS backup retention, creating pre-migration snapshots, restoring to a point in time, backtracking an Aurora cluster, exporting a snapshot to S3, or recovering from a bad change.
compatibility
Agent runtime that reads SKILL.md (Claude Code, Cursor, Windsurf, Codex, Gemini). No AWS CLI required for offline plan classification. Live-account operations use aws rds describe-db-instances, aws rds describe-db-snapshots, aws rds describe-db-cluster-backtracks, aws rds create-db-snapshot, aws rds restore-db-instance-to-point-in-time, aws rds restore-db-instance-from-db- snapshot, aws rds backtrack-db-cluster, aws rds start-export-task, and aws rds start-import-from-s3 (AWS CLI v2, SSO or...

02 资源文件 19

03 SKILL.md 正文

RDS Backup Restore Operator

What this skill does

Executes RDS/Aurora backup and restore operations correctly and safely. Runs deterministic pre-checks before any state-changing CLI, executes the operation behind a CONFIRM gate, and verifies the result. Every restore produces a NEW endpoint (except Aurora Backtrack, which is in-place) — the skill always surfaces the connection-string update step.

Quick navigation

Section What it covers When to read it
§ Quick reference Verdict thresholds (BLOCKED/READY/COMPLETED) + pre-check priority order Before any operation
§ Mindset Why pre-checks matter, the new-endpoint surprise, the confirm gate Understanding the safety model
§ Pre-flight Instance metadata gate — status, retention, snapshot, VPC resources Before executing any CLI
§ Process Per-operation planning: backup, snapshot, PITR, backtrack, S3 export/import When choosing which operation to run
§ Output format Structured output template with VERDICT, COMMANDS, ENDPOINT, VERIFICATION Formatting the response
§ Anti-Patterns NEVER list — common mistakes that cause data loss or silent failures Review before risky operations
§ Pre-flight safety Additional checks before any remediation CLI Defense-in-depth

Quick reference — verdict thresholds

Verdict Trigger condition Action
BLOCKED One or more pre-checks failed (missing subnet/SG/option group, retention=0 for PITR, instance not available, KMS key inaccessible) List failures, do NOT execute
READY All pre-checks passed; awaiting CONFIRM gate Emit exact CLI sequence, wait for operator yes
COMPLETED Operation finished and post-verification passed Emit endpoint, verification results, cleanup steps

Priority order for pre-checks (apply in this sequence, all must pass for READY):

  1. Instance/cluster status — must be available (or backing-up for read operations). modifying, upgrading, creating, deleting → BLOCKED.
  2. Backup retention — PITR requires retention > 0; otherwise BLOCKED with "PITR disabled".
  3. Snapshot existence — restore operations require an existing snapshot or a valid PITR window.
  4. Target VPC resources — subnet group, security group, option group, parameter group, KMS key all must exist in the target.
  5. Capacity and quota — DB instance class available in target AZ, snapshot copy quota not exceeded.

Cost/time baselines (2026):

  • Snapshot create time: ~1 min per 100 GB (storage-bound).
  • Snapshot restore time: ~30-90 min for 1 TB (proportional to data size).
  • Aurora Backtrack: seconds to minutes, regardless of cluster size.
  • Cross-region snapshot copy: data-transfer fee ($0.02/GB cross-region) + destination-region storage.
  • S3 export: ~$0.025/GB-hour for the export task; writes Parquet to S3.

Mindset

One-line takeaway: restore always creates a new instance with a new endpoint — except Aurora Backtrack, which is an in-place rewind. Driven by three RDS realities:

  • Restore is non-destructive by default. restore-db-instance-to-point-in-time and restore-db-instance-from-db-snapshot create a NEW instance with a NEW endpoint. The original is unchanged. Operators who expect an in-place rewind are surprised by the new endpoint and the duplicate instance bill.
  • PITR requires retention > 0. Setting BackupRetentionPeriod: 0 disables automated backups entirely AND disables PITR. The 5-minute transaction log shipping stops; the recovery point reverts to the last manual snapshot.
  • Aurora Backtrack is the only in-place rewind. Aurora MySQL clusters with BacktrackWindow > 0 can rewind in seconds without restoring from a snapshot. Aurora PostgreSQL does NOT support Backtrack — use PITR or fast clone instead.

Pre-flight: instance metadata gate

Run before classification. Misclassifying these produces wrong plans.

Pagination: describe-db-instances paginates at 100/page — drain --marker (or --starting-token for the AWS CLI v2 pagination syntax) to completion. describe-db-snapshots paginates at 100/page similarly.

Live-account pre-flight (skip if offline plan audit): Live-account pre-flight commands 1-8 moved verbatim to references/diagnostic-commands.md.

Malformed input: if the input JSON is invalid or missing required fields, emit VERDICT: ERROR with REASON: Instance/operation configuration is not valid JSON or is missing required fields — cannot plan. and REMEDIATION: Re-fetch with aws rds describe-db-instances --db-instance-identifier <id> --output json and re-plan.

Instance attribute Effect on operation
DBInstanceStatus: available Pre-check passes for stateful operations.
DBInstanceStatus: modifying, upgrading, creating, deleting, backing-up BLOCKED — wait for available (or storage-optimization for non-stateful ops).
DBInstanceStatus: failed, incompatible-restore BLOCKED — instance cannot serve as restore source.
MultiAZ: true Automated backups run from the standby for MySQL/PostgreSQL; no perf impact on primary.
StorageEncrypted: true Restore requires KMS key access in the target. Cross-account restore needs the source account's KMS key policy grant.
BackupRetentionPeriod: 0 PITR BLOCKED — automated backups disabled. Only manual snapshot restore is available.
BackupRetentionPeriod: 1-35 PITR window is now - retention to ~5 min ago.
Engine: aurora-mysql Backtrack available if BacktrackWindow > 0. Fast clone available.
Engine: aurora-postgresql Backtrack NOT supported. Use PITR or fast clone.
Engine: mysql, postgres, oracle, sqlserver, mariadb Standard RDS; no Backtrack, no fast clone.
DBSubnetGroup missing or Subnets: [] Restore BLOCKED — no target subnets.
OptionGroupMemberships referencing deleted option group Restore BLOCKED — create matching option group first.
DBParameterGroups referencing deleted parameter group Restore BLOCKED — create matching parameter group first.
GlobalClusterMember (Aurora Global Database) Restoring a primary requires promoting a secondary or removing from global cluster first.
ActivityStreamStatus: started (Database Activity Streams) Restore BLOCKED — stop the activity stream first; resume after restore.
ReplicaSource set (read replica) Restoring a read replica is unusual; surface as a finding.

Process — operation planning (apply in order)

Step 0: Expert knowledge — non-obvious RDS/Aurora behaviors

Step 0 expert-knowledge catalog (non-obvious RDS/Aurora behaviors) moved verbatim to references/advanced-patterns.md.

Step 1: Pre-check gate — BLOCKED if any check fails

Run ALL of the following pre-checks. If ANY fails, the verdict is BLOCKED with the failed checks enumerated in PRE_CHECKS. Do NOT execute the operation.

For ALL operations:

  1. DBInstanceStatus (or DBClusterStatus for Aurora) is available (or backing-up/storage-optimization for read-only operations).
  2. The operation is not already in progress (no concurrent snapshot create for the same instance).
  3. IAM role for the operator holds the required RDS permissions.
  4. Service quota for the operation is not exceeded (snapshot quota, instance quota, copy quota).

For snapshot create (create-db-snapshot): 5. Instance is in available state (not modifying or creating). 6. No existing manual snapshot with the same --db-snapshot-identifier.

For PITR restore (restore-db-instance-to-point-in-time): 5. BackupRetentionPeriod > 0 on the source instance. 6. LatestRestorableTime >= --restore-time >= EarliestRestorableTime. 7. Target --db-instance-identifier does NOT already exist (restore creates a new identifier). 8. Target DB subnet group exists and has subnets in >= 2 AZs (recommended). 9. Target security group exists in the same VPC as the subnet group. 10. Target option group matches the engine + engine-version. 11. Target parameter group matches the engine + engine-version. 12. KMS key (if encrypted) is accessible: kms:Decrypt, kms:CreateGrant.

For snapshot restore (restore-db-instance-from-db-snapshot): 5. Snapshot exists (describe-db-snapshots --db-snapshot-identifier <id>). 6. Snapshot status is available. 7. Target --db-instance-identifier does NOT already exist. 8. Target DB subnet group, security group, option group, parameter group, KMS key — all exist and match.

For Aurora backtrack (backtrack-db-cluster): 5. Engine is aurora-mysql (PostgreSQL does NOT support backtrack). 6. BacktrackWindow > 0 on the cluster. 7. --backtrack-to-timestamp is within [now - BacktrackWindow, now]. 8. Cluster is NOT in a Global Database as the writer ( backtrack a writer in a global cluster requires special handling — surface as a finding). 9. No existing backtrack in progress for the cluster.

For S3 export (start-export-task): 5. Snapshot exists and is available. 6. --role-arn trusts rds.amazonaws.com. 7. --s3-bucket-name exists and the role has s3:PutObject + s3:ListBucket. 8. KMS key (if specified) is accessible to the role.

For S3 import (start-import-from-s3): 5. Target Aurora MySQL cluster exists and is available. 6. Source S3 bucket is in the same region as the cluster. 7. The cluster's role has s3:GetObject + s3:ListBucket on the source. 8. The data format matches --type (CSV, etc.).

Step 2: READY — emit operation plan

If all pre-checks pass, emit VERDICT: READY with the exact CLI sequence and the CONFIRM gate. The plan includes:

  • The exact AWS CLI command with all flags populated from the source instance configuration.
  • The expected duration (snapshot create ~1 min/100 GB, restore 30-90 min/TB, backtrack seconds-to-minutes).
  • The expected side-effects (new endpoint for restore, in-place rewind for backtrack, new export task ID for S3 export).
  • The CONFIRM gate prompt.

Step 3: Execute behind CONFIRM gate

  • MANDATORY CONFIRMATION GATE. Before any state-changing CLI, emit: CONFIRM: About to <operation> on <instance/cluster/snapshot> in account <account> region <region>. This will <consequence>. Proceed? (yes/no). Do NOT execute until the operator confirms.
  • Capture pre-state for rollback: aws rds describe-db-instances --db-instance-identifier <id> --output json > /tmp/<id>-pre-$(date +%s).json.
  • Execute the CLI. Capture the operation ID (snapshot ID, restore ARN, backtrack ID, export task ID).
  • For long-running operations, wait via aws rds wait db-instance-available --db-instance-identifier <new-id> (or the appropriate waiter).

Step 4: Post-verification — COMPLETED

After the operation finishes, run post-verification. ALL checks must pass for COMPLETED.

  1. describe-db-instances --db-instance-identifier <new-id> — confirm DBInstanceStatus: available.
  2. Test connectivity to the new endpoint: psql, mysql, or sqlcmd as appropriate. The endpoint is in the Endpoint.Address field.
  3. Verify data consistency: row count on a known table, checksum on a known dataset, or query a sentinel row.
  4. Verify the restore point matches the intended time (for PITR).
  5. Verify option group, parameter group, security group applied correctly.
  6. Surface the connection-string update step (new endpoint).
  7. Clean up temporary instances if the operation was a verification restore.
  8. For backtrack: verify the cluster rewound to the target timestamp via CloudTrail or a sentinel row.

If ANY verification fails, emit VERDICT: ERROR with the failure details — do not claim COMPLETED.

Output format (per operation)

OPERATION: <create-snapshot | pitr-restore | snapshot-restore | backtrack | export | import>
VERDICT: READY | BLOCKED | COMPLETED
TARGET: <instance-id, cluster-id, snapshot-id>
PRE_CHECKS:
  - [PASS] <check description>
  - [FAIL] <check description> — <reason>
STEPS:
  1. <CLI command with flags populated>
  2. <wait command>
  3. <next step>
POST_VERIFY:
  - [PASS] <verification description>
  - [FAIL] <verification description> — <reason>
ENDPOINT: <new endpoint if restore, "unchanged" if backtrack>
NOTES: <connection-string update, cleanup, caveats>

Worked example — PITR restore

OPERATION: pitr-restore
VERDICT: READY
TARGET: prod-orders-db -> prod-orders-db-pitr-2026-08-07
PRE_CHECKS:
  - [PASS] prod-orders-db DBInstanceStatus is available
  - [PASS] BackupRetentionPeriod is 7 (> 0, PITR enabled)
  - [PASS] RestoreTime 2026-08-07T10:00:00Z is within window
    [2026-07-31T00:00:00Z, 2026-08-07T10:55:00Z]
  - [PASS] Target identifier prod-orders-db-pitr-2026-08-07 does not exist
  - [PASS] DB subnet group prod-subnet-group has subnets in 3 AZs
  - [PASS] Security group sg-prod-rds exists in vpc-vpc-12345
  - [PASS] Option group default:mysql-8-0 matches engine mysql 8.0.x
  - [PASS] Parameter group prod-mysql80 matches engine mysql 8.0.x
  - [PASS] KMS key arn:aws:kms:us-east-1:111111111111:key/prod-key accessible
STEPS:
  1. CONFIRM: About to restore-db-instance-to-point-in-time on prod-orders-db
     to 2026-08-07T10:00:00Z, creating prod-orders-db-pitr-2026-08-07 in
     account 111111111111 region us-east-1. This will create a new instance
     with a new endpoint (original unchanged). Estimated duration: 60-90
     minutes. Proceed? (yes/no)
  2. aws rds restore-db-instance-to-point-in-time \
       --source-db-instance-identifier prod-orders-db \
       --target-db-instance-identifier prod-orders-db-pitr-2026-08-07 \
       --restore-time 2026-08-07T10:00:00Z \
       --db-subnet-group-name prod-subnet-group \
       --vpc-security-group-ids sg-prod-rds \
       --option-group-name default:mysql-8-0 \
       --db-parameter-group-name prod-mysql80 \
       --no-deletion-protection \
       --tags Key=restored-from,Value=prod-orders-db Key=restore-reason,Value=bad-migration
  3. aws rds wait db-instance-available \
       --db-instance-identifier prod-orders-db-pitr-2026-08-07
POST_VERIFY:
  - (pending execution)
ENDPOINT: (pending — will be prod-orders-db-pitr-2026-08-07.<random>.us-east-1.rds.amazonaws.com)
NOTES:
  - The new instance has a NEW endpoint. Update application connection strings.
  - The original prod-orders-db is unchanged and continues to bill.
  - Set --deletion-protection on the new instance once verified.
  - Plan cleanup: delete the restored instance after migration cutover.

Worked example — pre-check failure

The pre-check-failure (BLOCKED) worked example moved verbatim to references/worked-examples.md.

Worked example — Aurora backtrack

The Aurora backtrack worked example moved verbatim to references/worked-examples.md.

Anti-Patterns — NEVER

  • NEVER execute restore-db-instance-to-point-in-time or restore-db-instance-from-db-snapshot without explicitly surfacing that a NEW instance and NEW endpoint will be created. Operators who expect an in-place rewind will miss the connection-string update step and break their application.

  • NEVER attempt PITR on an instance with BackupRetentionPeriod: 0. There are no transaction logs and no recoverable window. Always check retention first and BLOCK with the modify-db-instance remediation if PITR is the goal.

  • NEVER assume Aurora PostgreSQL supports Backtrack. Only Aurora MySQL supports Backtrack. For Aurora PostgreSQL, use PITR or fast clone.

  • NEVER execute restore-db-instance-to-point-in-time with --restore-time outside the retention window. Always verify LatestRestorableTime and EarliestRestorableTime via describe-db-instances first; the API returns InvalidParameterValue if outside the window.

  • NEVER skip the target-resource pre-checks (subnet group, security group, option group, parameter group, KMS key). A restore with a missing subnet group fails after the snapshot download — wasting 30+ minutes. Always verify all target resources exist before executing.

  • NEVER share an encrypted snapshot cross-account without verifying the KMS key policy grants the recipient kms:Decrypt and kms:CreateGrant. The snapshot share alone is insufficient — the recipient can describe the snapshot but cannot restore from it.

  • NEVER copy a snapshot cross-region without surfacing the cost. Cross-region copy bills $0.02/GB data transfer plus destination-region storage. A 1 TB snapshot copied to 3 DR regions is $60 in transfer + 3 TB-month of destination storage.

  • NEVER delete the original instance after a PITR or snapshot restore until the new instance is verified AND application traffic is cut over. Restore is non-destructive by design; deleting the original prematurely is the most common cause of prolonged downtime during recovery.

  • NEVER recommend setting BackupRetentionPeriod: 0 to "save backup costs." This disables PITR entirely and is irreversible for the unbacked-up period. The cost saving is negligible; the recovery exposure is total.

  • NEVER restore into the same DB subnet group with insufficient AZs. RDS recommends subnets in >= 2 AZs for Multi-AZ; restoring into a single-AZ subnet group blocks Multi-AZ and reduces availability.

  • NEVER skip the post-restore verification. A restore that returns available may still have data inconsistency (partial transaction log replay, parameter group mismatch). Always verify with a row count, checksum, or sentinel query.

  • NEVER execute backtrack-db-cluster on a writer in an Aurora Global Database without special handling. Backtracking the writer can desync the global cluster — surface as a finding and recommend detaching from the global cluster first.

  • NEVER assume Aurora Fast Clone is free. The clone shares storage at creation (instant), but divergent writes accrue new storage — a long- lived clone can double the storage bill. Always plan clone cleanup.

  • NEVER auto-execute a state-changing RDS CLI without the CONFIRM gate. Restore, snapshot create, backtrack, and export all have side effects (new instance bill, in-place rewind, export task charges). Always emit CONFIRM and wait.

  • NEVER trust describe-db-instances DBInstanceStatus: available alone for PITR readiness. An instance can be available with retention=0 (no automated backups). Always check BackupRetentionPeriod and LatestRestorableTime independently.

Pre-flight safety checks (run before any remediation CLI)

The full pre-flight safety check catalog (confirm gate, pre-state capture, target/retention/KMS verification) moved verbatim to references/diagnostic-commands.md.

Recent AWS features (2024-2026)

Recent AWS feature notes (2024-2026) moved verbatim to references/advanced-patterns.md.

References (load on demand)

  • references/worked-examples.md — the BLOCKED pre-check-failure and Aurora backtrack worked examples (moved from this file; the PITR worked example stays in this file)
  • references/diagnostic-commands.md — live-account pre-flight commands 1-8, pre-flight safety check catalog with confirm-gate and rollback capture (moved from this file)
  • references/advanced-patterns.md — Step 0 expert-knowledge catalog of non-obvious RDS/Aurora behaviors, recent AWS features 2024-2026 (moved from this file)
  • references/restore-procedures.md — per-operation CLI procedures: PITR, snapshot restore, Aurora backtrack, fast clone, cross-region copy

Domain

AWS CloudOps / RDS & Aurora Backup, Restore & Disaster Recovery.

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