[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M7AB5EJPCKRAEBXH97588VAP":3,"skill-same:01M7AB5EJPCKRAEBXH97588VAP":144},{"skillNo":4,"name":5,"nameZh":6,"description":7,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":16,"categories":17,"keywords":18,"publishedAt":19,"linkOnly":20,"markdownContent":21,"markdownBody":22,"frontMatter":23,"refName":40,"commitSha":41,"skillDirectory":42,"forksCount":43,"repositoryHtmlUrl":44,"repositoryHomepageUrl":6,"ownerHtmlUrl":45,"ownerBlogUrl":6,"ownerDescription":46,"updatedAt":47,"install":48,"resources":61},"01M7AB5EJPCKRAEBXH97588VAP","redshift-cluster-auditor",null,"Audits Amazon Redshift provisioned clusters for security posture and configuration gaps — public accessibility (internet-exposed cluster), KMS encryption-at-rest (immutable post-creation), require_ssl parameter group enforcement, S3 audit logging, VPC security-group ingress on the cluster port, automated-snapshot retention (PITR), and enhanced VPC routing (COPY\u002FUNLOAD traffic path). Emits a deterministic categorical verdict (PUBLIC | NO_ENCRYPTION | NO_SSL | NO_AUDIT_LOG | CONFIG_GAP | OK) per cluster with enumerated findings and specific remediation. Use when reviewing a Redshift cluster before production deployment, auditing encryption or SSL posture, checking S3 audit logging coverage, validating snapshot retention, or hardening data-warehouse security.","jhkchan\u002Faws-agent-skills","jhkchan","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F1083902?v=4","01M7A2SVQYP9GQTPTDXP5JKYQ4","Jacky Chan","01M7A2SVS117YWCTZABB75HZR1",4,"Apache-2.0","https:\u002F\u002Fgithub.com\u002Fjhkchan\u002Faws-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Fredshift-cluster-auditor\u002FSKILL.md",[],[],"2026-11-15 14:19:17",false,"---\nname: redshift-cluster-auditor\ndescription: Audits Amazon Redshift provisioned clusters for security posture and configuration gaps — public accessibility (internet-exposed cluster), KMS encryption-at-rest (immutable post-creation), require_ssl parameter group enforcement, S3 audit logging, VPC security-group ingress on the cluster port, automated-snapshot retention (PITR), and enhanced VPC routing (COPY\u002FUNLOAD traffic path). Emits a deterministic categorical verdict (PUBLIC | NO_ENCRYPTION | NO_SSL | NO_AUDIT_LOG | CONFIG_GAP | OK) per cluster with enumerated findings and specific remediation. Use when reviewing a Redshift cluster before production deployment, auditing encryption or SSL posture, checking S3 audit logging coverage, validating snapshot retention, or hardening data-warehouse security.\nlicense: Apache-2.0\ncompatibility: Agent runtime that reads SKILL.md (Claude Code, Cursor, Windsurf, Codex, Gemini). No AWS CLI required for offline metadata classification. Live-account audits use aws redshift describe-clusters, describe-logging-status, describe-cluster-parameter-groups (with describe-cluster-parameters for the PG), aws ec2 describe-security-groups, and aws kms describe-key (AWS CLI v2, SSO or key-based credentials).\nmetadata:\n  domain: aws-cloudops\n  complexity: high\n  requires_llm: 'true'\n  phase: '2'\n  supports_pipeline: 'true'\n  entry_point: 'false'\n  family: Analytics\n  verdict_shape: PUBLIC | NO_ENCRYPTION | NO_SSL | NO_AUDIT_LOG | CONFIG_GAP | OK\n  when_to_use: Reviewing a Redshift provisioned cluster before production deployment, auditing encryption-at-rest, validating require_ssl enforcement, checking S3 audit logging coverage, inspecting automated-snapshot retention, evaluating enhanced VPC routing posture, or hardening a data warehouse for a compliance review.\n  activation_triggers: audit this Redshift cluster, is my Redshift cluster public, is encryption enabled on Redshift, is require_ssl on, is Redshift audit logging configured, are automated snapshots enabled, enhanced VPC routing Redshift, Redshift security group audit, harden this data warehouse\n  invocation_schema: 'Input: either (a) a describe-clusters Cluster block, optionally paired with describe-logging-status, describe-cluster-parameters (for the attached parameter group), and describe-security-groups outputs, OR (b) a ClusterIdentifier for live-account audit. Output: deterministic CLUSTER\u002FVERDICT\u002FREASON\u002FFINDINGS\u002FREMEDIATION block per cluster, where VERDICT ∈ {PUBLIC, NO_ENCRYPTION, NO_SSL, NO_AUDIT_LOG, CONFIG_GAP, OK, ERROR}.'\n  version: 0.1.0\n  author: Jacky Chan — AWS Community Builder\n  keywords: Redshift, data warehouse, PubliclyAccessible, encryption, KMS, require_ssl, SSL, TLS, parameter group, audit logging, S3 audit log, enable_user_activity_logging, automated snapshots, AutomatedSnapshotRetentionPeriod, enhanced VPC routing, EnhancedVPCRouting, COPY, UNLOAD, VPC security group, cluster port 5439, data-warehouse hardening, compliance, Redshift Serverless\n  tags: redshift, analytics, security, encryption, ssl, audit-logging, snapshots, vpc, compliance, audit\n---\n\n# Redshift Cluster Auditor\n\n## Mindset\n\n**One-line takeaway:** the verdict is always the **worst** finding across\nsix ordered dimensions, and `PubliclyAccessible: true` on a Redshift\ncluster is in a class of its own — it places a petabyte-scale data\nwarehouse on the public internet with no VPC boundary in front of it.\n\nRedshift clusters concentrate an organisation's most sensitive analytics\ndata: customer records, financial transactions, product analytics. Three\nRedshift-specific behaviours drive the classification:\n\n- **`PubliclyAccessible: true` is a direct internet endpoint.** Unlike RDS\n  (where the SG + PubliclyAccessible + subnet routing combine), Redshift\n  with `PubliclyAccessible: true` is reachable from `0.0.0.0\u002F0` over TCP\n  5439 (or 5440). There is no \"public subnet routing\" subtlety — the flag\n  is the exposure. A public Redshift cluster is a public data warehouse.\n- **Encryption-at-rest is immutable per cluster.** Unlike RDS or EBS,\n  there is no online modify flow and no snapshot-restore toggle. The\n  *only* path is: provision a new encrypted cluster, `UNLOAD` from the\n  old, `COPY` into the new, redirect downstream BI tools, decommission\n  the old. Treat `Encrypted: false` as a migration backlog, not a CLI\n  toggle.\n- **`require_ssl` lives in the parameter group, not the cluster metadata.**\n  The default value is `false` and the default parameter group\n  (`default.redshift-1.0`) cannot be modified — a cluster on the default\n  PG is reporting the engine default, not an operator decision. The\n  parameter must be set to `true` in a *custom* parameter group, and the\n  cluster must be rebooted for the change to apply.\n\n## Quick reference — verdict priority\n\n| Condition | Verdict | Step |\n|---|---|---|\n| `PubliclyAccessible: true` | **PUBLIC** | Step 1 |\n| `Encrypted: false` | **NO_ENCRYPTION** | Step 2 |\n| `require_ssl` parameter value is `false` (or unset) | **NO_SSL** | Step 3 |\n| `LoggingEnabled: false` (describe-logging-status) | **NO_AUDIT_LOG** | Step 4 |\n| Any of: snapshot retention 0, enhanced VPC routing off, SG ingress `0.0.0.0\u002F0` on cluster port | **CONFIG_GAP** | Step 5 |\n| All dimensions pass | **OK** | Step 6 |\n\nApply in order; the first matching verdict wins. The ordering is\ndeliberate: PUBLIC > NO_ENCRYPTION > NO_SSL > NO_AUDIT_LOG > CONFIG_GAP > OK.\nAll findings still appear in the FINDINGS list regardless of which one\nbecame the verdict. See the ordered steps below for Redshift-specific edge\ncases (AWS-managed vs customer-managed KMS, default vs custom parameter\ngroup, Serverless deferral, HSM legacy clusters).\n\n## Pre-flight: cluster metadata gate (run before classification)\n\nBefore evaluating any dimension, classify the cluster itself. Several\ncluster attributes **short-circuit** the audit — misclassifying them\nproduces false positives that erode operator trust.\n\n**Live-account pre-flight checks (skip if doing offline metadata audit):**\n\n1. **Confirm the cluster exists and is reachable.** `aws redshift\n   describe-clusters --cluster-identifier \u003Cid> --profile \u003Cp>` — fail\n   closed (skip the audit, surface an ERROR) if it returns\n   `ClusterNotFound`. A cluster in `deleting` or `final-snapshot` state\n   cannot be remediated in place; note the lifecycle state in the\n   verdict.\n2. **Verify the caller can run `redshift:Describe*` AND\n   `ec2:DescribeSecurityGroups`.** The SG dimension requires\n   cross-service access — a read-only auditor role scoped to Redshift\n   only will silently emit CONFIG_GAP false positives because it cannot\n   see the SG rules. Verify BEFORE the audit, not after.\n3. **Snapshot the cluster's parameter group application.** Reboot\n   windows are required for parameter-group changes that affect SSL.\n   Confirm whether a recent reboot is in flight\n   (`ClusterParameterGroupStatus: Applying` means pending reboot).\n4. **Confirm whether audit logging was ever enabled then disabled.**\n   `describe-logging-status` returns `LoggingEnabled: false` for both\n   \"never enabled\" and \"previously enabled, now off\". For forensics,\n   check the bucket for prior log objects under the cluster's prefix.\n\n| Attribute | Value | Effect on audit |\n|---|---|---|\n| `ClusterStatus` | `available`, `modifying`, `rebooting` | Proceed normally. `modifying`\u002F`rebooting` may mean a parameter-group change is mid-apply. |\n| `ClusterStatus` | `deleting`, `final-snapshot` | Cluster is being decommissioned — emit verdict with a note that remediation is N\u002FA until the cluster is recreated. |\n| `ClusterStatus` | `paused` (RA3 limit, rare) | Cluster is administratively paused. Encryption \u002F SSL \u002F logging dimensions still audit. |\n| `NumberOfNodes: 1` | Single-node | Single-node clusters do **not** provide continuous availability during node failure. Note as an operational risk (not a verdict driver). Multi-AZ deployment (the `MultiAZ` flag on supported node types) is the availability dimension — separate from security. |\n| `NodeType: serverless` or input is a Redshift Serverless workgroup | **Skip this skill.** Redshift Serverless has a different API surface (`redshift-serverless get-workgroup`, no `PubliclyAccessible`, no `ClusterParameterGroupName` — SSL is `require_ssl` on the config). Defer to a Serverless-specific audit. Emit ERROR with a routing note. |\n| `HsmClientCertificateIdentifier` present | **Legacy HSM-encrypted cluster.** Pre-2017 Redshift supported on-prem HSM encryption. HSM-managed clusters report `Encrypted: true` but bypass KMS entirely. Treat as encrypted for verdict purposes; flag as a deprecation risk (HSM integration is end-of-life). |\n| `ElasticIpStatus: ElasticIp` present | **EIP attached.** A public IP pinned to the cluster. Combined with `PubliclyAccessible: true`, this is a stable, DNS-portable public database endpoint — worse than an ephemeral public IP because it survives stop\u002Fstart. |\n| `KmsKeyId: null` AND `Encrypted: true` | **AWS-managed KMS key** (`aws\u002Fredshift`). Encryption is on; you do NOT control rotation or cross-account policy. Note in findings — CMK migration is a hardening step, not a security gap. |\n| `KmsKeyId: \u003Carn>` AND `Encrypted: true` | **Customer-managed KMS key.** Preferred posture. Cross-reference the key policy via the kms-key-policy-auditor skill. |\n| `ClusterParameterGroupName: default.redshift-1.0` (or `.0` for any engine version) | **Default parameter group.** Cannot be modified — `require_ssl` and `enable_user_activity_logging` cannot be set to `true` on the default PG. Any non-default parameter value requires migrating to a custom PG. |\n\n**If the cluster metadata is malformed** (missing `ClusterIdentifier`,\nmissing `PubliclyAccessible`, missing `Encrypted`), output:\n\n```text\nCLUSTER: \u003Cid-or-unknown>\nVERDICT: ERROR\nREASON: Cluster metadata is incomplete — cannot classify (missing \u003Cfield>).\nREMEDIATION: Re-fetch with `aws redshift describe-clusters --cluster-identifier \u003Cid> --output json` and re-audit.\n```\n\n## Process — Classification logic (apply in order, first match wins)\n\n### Step 0: Expert knowledge — non-obvious Redshift behaviours that change classification\n\nFourteen non-obvious Redshift behaviours (PubliclyAccessible vs EnhancedVPCRouting, require_ssl scope, read-only default parameter group, AWS-managed key, encryption immutability, snapshot retention expiry and range, enable-logging vs enable_user_activity_logging, port 5439, cross-region snapshot copy, paused clusters, single-node exposure, CloudTrail control-plane only): moved verbatim to [references\u002Fadvanced-patterns.md](references\u002Fadvanced-patterns.md).\nLoad before classification; Steps 1-6 below assume these constraints.\n\n### Step 1: Public accessibility (highest priority — internet-exposed data warehouse)\n\nIf `PubliclyAccessible: true`:\n\n- **Verdict: PUBLIC.** The cluster is assigned a public IP and accepts\n  connections from `0.0.0.0\u002F0` on the cluster port (subject to SG\n  rules). Even if the SG restricts ingress to a corporate CIDR, the\n  cluster responds to public TCP probes (SYN-ACK) — it is discoverable\n  by any internet scanner, and any future SG widening instantly exposes\n  it. A public Redshift cluster is a public petabyte-scale database.\n- The presence of an Elastic IP (`ElasticIpStatus.ElasticIp` set) makes\n  the exposure *stable* — the public IP survives stop\u002Fstart. Note this\n  in FINDINGS; it does not change the verdict but raises the\n  operational risk.\n\nPubliclyAccessible is evaluated first because it is the only dimension\nwhere the cluster is **directly attackable from anywhere on the\ninternet**. Encryption, SSL, and logging are all important but\npresuppose the cluster is reachable; PUBLIC short-circuits that\nassumption.\n\n### Step 2: Encryption-at-rest (immutable post-creation)\n\nIf `Encrypted: false`:\n\n- **Verdict: NO_ENCRYPTION.** Data-at-rest is unencrypted. Snapshots\n  inherit the cluster's encryption state (an unencrypted cluster\n  produces unencrypted snapshots; you can copy a snapshot with\n  encryption enabled, which is one path to a migration). The KMS key\n  flag is irrelevant when `Encrypted: false`.\n\nIf `Encrypted: true`:\n- With `KmsKeyId: null` → AWS-managed key (`aws\u002Fredshift`). OK for this\n  dimension. Note as an additive finding if CMK is required by the\n  compliance framework.\n- With `KmsKeyId: \u003Carn>` → customer-managed key. OK for this dimension.\n  Recommend cross-referencing the key policy via the\n  kms-key-policy-auditor skill (the key's policy is a separate audit\n  surface).\n\nEncryption is evaluated second because it is the second-highest blast\nradius (after internet exposure): if an unencrypted snapshot leaks or\nthe underlying storage is improperly decommissioned, the data is\nplaintext.\n\n### Step 3: require_ssl parameter group enforcement\n\nInspect the parameter group attached to the cluster\n(`ClusterParameterGroupName`) and its parameter list. Find the\n`require_ssl` parameter.\n\n- **`require_ssl: false` (any source)** → **Verdict: NO_SSL.** SQL\n  clients MAY connect over plaintext. Compliance frameworks (PCI-DSS\n  4.1, HIPAA Transport Encryption, SOC 2 CC6.1) require TLS in transit.\n  Redshift's wire protocol supports SSL; `require_ssl: true` enforces\n  it.\n\n- **`require_ssl: true`** → OK for this dimension.\n- **`require_ssl` parameter absent from the response** → treat as\n  `false` (the engine default is false). Flag as NO_SSL.\n\n**Default parameter group trap:** if the cluster is on\n`default.redshift-1.0` and `Source: engine-default`, the operator has\nnot actively decided. Remediation requires migrating to a custom PG —\nemphasise this in REMEDIATION. A custom PG with `require_ssl: false`\nand `Source: user` is an explicit operator decision to allow plaintext\n(a deliberate misconfiguration, not a default).\n\n**Pending-reboot trap:** if `ClusterParameterGroupStatus: Applying` or\n`pending-reboot`, a recent parameter change has not yet taken effect.\nThe current effective value is the OLD value, not the new one. Surface\nthe pending value in the FINDINGS; the verdict reflects the current\neffective state.\n\n### Step 4: Audit logging (S3 export)\n\nInspect `describe-logging-status`. The key field is `LoggingEnabled`.\n\n- **`LoggingEnabled: false`** → **Verdict: NO_AUDIT_LOG.** Connection,\n  user-activity, and DDL events are not being exported to S3. This\n  eliminates forensic capability for \"who ran what query when\" and\n  breaks compliance evidence chains (PCI-DSS 10.2, HIPAA §164.312(b),\n  SOC 2 CC7.2).\n\n- **`LoggingEnabled: true`** → OK for this dimension. Note the bucket\n  name and prefix in FINDINGS; recommend verifying the bucket's\n  object-lock \u002F retention policy and access controls separately.\n\nAudit logging is evaluated fourth because it is the forensic\nprerequisite for investigating the other findings. Without it, a\nPUBLIC+NO_ENCRYPTION breach cannot be reconstructed.\n\n### Step 5: Configuration gaps (additive findings)\n\nEvaluate the remaining dimensions and surface any failures as CONFIG_GAP\nfindings. If any of the following fail AND Steps 1-4 all passed, the\nverdict is **CONFIG_GAP**.\n\n| Sub-check | Failure condition | Why it matters |\n|---|---|---|\n| **Automated snapshots** | `AutomatedSnapshotRetentionPeriod: 0` | PITR disabled. Existing automated snapshots expire within the next snapshot window. Manual snapshots are unaffected but require active management. |\n| **Enhanced VPC routing** | `EnhancedVPCRouting: false` | `COPY` \u002F `UNLOAD` traffic bypasses VPC SGs, NACLs, and VPC endpoints — data path is not subject to network controls. Engine default is `false`. |\n| **VPC security group ingress on cluster port** | Any SG rule with `0.0.0.0\u002F0` (or `::\u002F0`) on the cluster's TCP port (default 5439, configurable via `Endpoint.Port`) | Even with `PubliclyAccessible: false`, an `0.0.0.0\u002F0` SG rule on the cluster port is a CONFIG_GAP — a future flip of `PubliclyAccessible` to `true` instantly exposes the cluster. Note: SG ingress does NOT make a private cluster public; it expands the in-VPC attack surface. |\n| **User-activity logging** | `enable_user_activity_logging: false` parameter value | STL_QUERY retains query text for ~3-7 days; without user_activity_logging exported to S3 (via the audit log pipeline), long-horizon forensic queries are impossible. Treat as CONFIG_GAP (additive to Step 4 — if Step 4 failed, this is folded into NO_AUDIT_LOG). |\n| **Snapshot retention below compliance floor** | `AutomatedSnapshotRetentionPeriod: 1` on a production cluster | The engine default is 1 day. PCI\u002FHIPAA typically require 7+; flag as a soft CONFIG_GAP if compliance framework applies. |\n| **Cross-region snapshot copy absent** | `ClusterSnapshotCopyStatus` is absent on a production cluster | Single-region snapshots = no DR. Flag as a resilience CONFIG_GAP (not security). |\n\nFor the SG check specifically: the SG dimension surfaces the *exposure*\nof the cluster within its VPC. SG rules from corporate CIDRs\n(`10.0.0.0\u002F8`, `172.16.0.0\u002F12`, `192.168.0.0\u002F16`) are typically fine;\n`0.0.0.0\u002F0` on the cluster port is the canonical misconfiguration.\n\n### Step 6: Aggregation — OK verdict\n\nIf all dimensions pass (PUBLIC not set, Encrypted true, require_ssl true,\nLoggingEnabled true, no CONFIG_GAP sub-findings), the verdict is **OK**.\n\n```text\nverdict_priority = PUBLIC > NO_ENCRYPTION > NO_SSL > NO_AUDIT_LOG > CONFIG_GAP > OK\n```\n\nApply in order; the first matching step wins. ALL findings (including\nnon-verdict-driving ones) appear in the FINDINGS list.\n\n## Output format (per cluster)\n\n```text\nCLUSTER: \u003Ccluster-identifier>\nVERDICT: PUBLIC | NO_ENCRYPTION | NO_SSL | NO_AUDIT_LOG | CONFIG_GAP | OK\nREASON: \u003C1-2 sentences citing the worst finding and its step number>\nFINDINGS:\n  - [PUBLIC] \u003Cfinding description (Step 1)>\n  - [NO_SSL] \u003Cfinding description (Step 3)>\n  - [OK] \u003Cdimension that passed>\nREMEDIATION: \u003Cspecific action per finding, or \"None required\" if OK>\n```\n\n### Worked example — public cluster with encryption disabled\n\n```text\nCLUSTER: public-cluster-no-encryption\nVERDICT: PUBLIC\nREASON: PubliclyAccessible is true, placing the cluster on the public\ninternet on TCP 5439 (Step 1). Encryption is also disabled, require_ssl\nis false, and S3 audit logging is off — every dimension fails.\nFINDINGS:\n  - [PUBLIC] PubliclyAccessible: true — cluster accepts connections from\n    0.0.0.0\u002F0 on the cluster port (Step 1)\n  - [NO_ENCRYPTION] Encrypted: false — data-at-rest is plaintext (Step 2)\n  - [NO_SSL] require_ssl: false in default.redshift-1.0 — clients may\n    connect over plaintext (Step 3)\n  - [NO_AUDIT_LOG] LoggingEnabled: false — no S3 audit trail (Step 4)\n  - [CONFIG_GAP] EnhancedVPCRouting: false — COPY\u002FUNLOAD bypasses VPC\n    controls (Step 5)\n  - [CONFIG_GAP] SG sg-0aaa11112222 has 0.0.0.0\u002F0 on TCP 5439 (Step 5)\nREMEDIATION:\n  1. PUBLIC — Set PubliclyAccessible to false: see CLI below. This\n     requires a client reboot. If public access is INTENTIONAL (e.g., a\n     partner-data workload), restrict the SG to specific partner CIDRs\n     immediately.\n  2. NO_ENCRYPTION — Plan an encryption migration: provision a new\n     encrypted cluster, UNLOAD to S3, COPY into the new cluster, redirect\n     downstream tools, decommission the old cluster. Encryption cannot be\n     toggled on an existing cluster.\n  3. NO_SSL — Create a custom parameter group with require_ssl: true,\n     associate it with the cluster, and reboot for the change to take\n     effect.\n  4. NO_AUDIT_LOG — Enable S3 audit logging:\n     aws redshift enable-logging --cluster-identifier public-cluster-no-encryption \\\n       --log-destination-name S3 --bucket-name audit-logs-111111111111 \\\n       --s3-key-prefix redshift\u002Fpublic-cluster-no-encryption\u002F\n  5. CONFIG_GAP — Enable EnhancedVPCRouting (requires cluster reboot) and\n     restrict the SG ingress to known CIDRs on the cluster port.\n```\n\n## Edge-case handling\n\nNine edge cases (modifying\u002Frebooting states, default parameter group, Multi-AZ, RA3\u002FDC2\u002FDS2 node types, HSM legacy, snapshot copy status, Serverless deferral, AWS-managed key, non-default port): moved verbatim to [references\u002Fadvanced-patterns.md](references\u002Fadvanced-patterns.md).\nConsult after the ordered steps when the cluster metadata is unusual.\n\n## Anti-Patterns — NEVER\n\n- NEVER classify `PubliclyAccessible: true` as anything other than\n  PUBLIC. The flag is a direct internet endpoint on TCP 5439\u002F5440. SG\n  rules may restrict ingress, but the cluster is still discoverable and\n  attackable by any internet scanner. A public data warehouse is the\n  highest-impact Redshift misconfiguration.\n\n- NEVER recommend \"just enable encryption\" on an existing cluster.\n  Redshift encryption is immutable per cluster. The only remediation\n  is a new cluster + UNLOAD\u002FCOPY migration. Stating otherwise causes\n  operators to attempt `modify-cluster --encrypted` (which does not\n  exist) and lose trust in the audit.\n\n- NEVER treat `require_ssl: false` from the default parameter group as\n  an operator decision. `Source: engine-default` means the operator\n  has not hardened this — they may not even know the parameter exists.\n  Remediation requires migrating to a custom PG (the default PG is\n  read-only).\n\n- NEVER assume `enable-logging` and `enable_user_activity_logging` are\n  the same thing. `enable-logging` (the API) exports audit events to\n  S3. `enable_user_activity_logging` (the parameter) logs every SQL\n  query to STL_QUERY. They are separate controls with separate\n  forensic value. Treat absence of either as a gap.\n\n- NEVER assume CloudTrail covers Redshift SQL activity. CloudTrail logs\n  control-plane events (CreateCluster, ModifyCluster, DeleteCluster).\n  It does NOT log data-plane SQL queries — that is exclusively the\n  domain of `enable-logging` S3 audit logs. \"We have CloudTrail\" does\n  not satisfy \"we audit database activity\" for Redshift.\n\n- NEVER conflate `PubliclyAccessible: false` with \"no internet exposure\n  for data transfers\". `EnhancedVPCRouting: false` (the default) means\n  COPY\u002FUNLOAD traffic to S3 traverses the public AWS network path,\n  bypassing your VPC's SGs, NACLs, and VPC endpoints. Privacy of the\n  cluster endpoint and privacy of the bulk-transfer data path are\n  independent concerns.\n\n- NEVER treat a single-node cluster as having high availability.\n  Single-node clusters have no replication and no failover. Destructive\n  queries (DROP TABLE, TRUNCATE) affect the entire dataset with no\n  redundant copy. Note this as an operational risk; do not let it\n  drive the security verdict.\n\n- NEVER flag `KmsKeyId: null` with `Encrypted: true` as an encryption\n  gap. This is the AWS-managed key (`aws\u002Fredshift`) — encryption is\n  on. Rotates annually on AWS's schedule. Acceptable for most\n  compliance frameworks. Flag CMK as a hardening opportunity, not a\n  finding.\n\n- NEVER attempt to apply this skill to a Redshift Serverless workgroup.\n  Serverless has a different API surface (no `PubliclyAccessible`, no\n  `ClusterParameterGroupName`, SSL enforced via `config-parameters` on\n  the workgroup). Emit ERROR with a routing note — misclassifying a\n  Serverless workgroup as PUBLIC or NO_SSL is a false positive.\n\n- NEVER assume `AutomatedSnapshotRetentionPeriod: 0` only disables\n  future snapshots. Redshift immediately begins expiring existing\n  automated snapshots. Operators setting retention to 0 to \"save cost\"\n  without first converting critical snapshots to manual lose PITR\n  irreversibly within hours.\n\n- NEVER recommend modifying a default parameter group. The default PG\n  (`default.redshift-1.0`) is read-only — `modify-cluster-parameter-group`\n  fails. Remediation requires creating a custom PG and associating it\n  with the cluster.\n\n- NEVER treat an SG with `0.0.0.0\u002F0` on TCP 5439 as making the cluster\n  public. The cluster is public ONLY if `PubliclyAccessible: true`. SG\n  `0.0.0.0\u002F0` ingress on a private cluster is a CONFIG_GAP — it\n  expands the in-VPC attack surface and sets up a future exposure if\n  the flag is flipped, but it does not by itself expose the cluster to\n  the internet.\n\n- NEVER evaluate the SG ingress on the wrong port. The cluster port is\n  in `Endpoint.Port` (default 5439, customisable up to 65535). SG\n  rules matching TCP 5432 (Postgres default) silently fail to permit\n  traffic. Read the actual port from the metadata, not the default.\n\n- NEVER flag a cluster in `paused` state as safer or out-of-scope.\n  Paused clusters retain their encryption, SG, and logging posture;\n  the audit is identical. Pause is a cost-optimisation, not a security\n  control.\n\n## Pre-flight safety checks (run before any remediation CLI)\n\nSeven pre-flight safety checks (CONFIRM gate, reboot warnings, encryption migration scope, pre-change snapshot, audit bucket ownership, cross-region cost, shared-SG verification): moved verbatim to [references\u002Fdiagnostic-commands.md](references\u002Fdiagnostic-commands.md).\nDefense-in-depth before any remediation CLI.\n\n## Remediation guidance\n\n**Remediation ordering principle:** additive changes first (enable\nlogging, associate a new PG), destructive \u002F state-changing changes last\n(modify cluster, reboot). Public accessibility and encryption are the\ntwo findings that cannot be silently fixed in place — plan migrations,\nnot toggles.\n\nPer-verdict remediation playbooks with full CLI blocks (PUBLIC, NO_ENCRYPTION UNLOAD\u002FCOPY migration, NO_SSL custom parameter group, NO_AUDIT_LOG enable-logging, CONFIG_GAP sub-findings, OK follow-ups): moved verbatim to [references\u002Fremediation-procedures.md](references\u002Fremediation-procedures.md).\nEmit per-finding remediation from that reference.\n\n## Recent AWS features (2024-2026)\n\nFour recent AWS features 2024-2026 (Serverless enhancements, Aurora Zero-ETL, data sharing, Redshift ML): moved verbatim to [references\u002Fadvanced-patterns.md](references\u002Fadvanced-patterns.md).\nConsult before citing feature limits or recency.\n\n## Related skills\n\n- **kms-key-policy-auditor:** audit the KMS key policy when a\n  customer-managed key encrypts the cluster. The cluster audit treats\n  the key as a black box; the key-policy audit determines who can\n  decrypt cluster snapshots.\n- **ec2-security-group-auditor:** deeper SG analysis including\n  prefix-list composition and cross-VPC exposure (the cluster audit\n  checks only the cluster port's exposure on the attached SGs).\n- **s3-public-access-auditor:** audit the S3 bucket receiving UNLOAD\n  exports or audit logs — a private cluster with a public UNLOAD bucket\n  silently exfiltrates data.\n\n## References (load on demand)\n\n- [references\u002Fadvanced-patterns.md](references\u002Fadvanced-patterns.md) — Step 0 non-obvious behaviours, edge-case handling, recent AWS features (2024-2026).\n- [references\u002Fdiagnostic-commands.md](references\u002Fdiagnostic-commands.md) — Pre-flight safety checks before any remediation CLI.\n- [references\u002Fremediation-procedures.md](references\u002Fremediation-procedures.md) — Per-verdict remediation playbooks with CLI blocks.\n\n## Domain\n\nAWS CloudOps \u002F Redshift Analytics Security & Compliance.\n\n## AWS documentation\n\n- **Amazon Redshift documentation** — https:\u002F\u002Fdocs.aws.amazon.com\u002Fredshift\u002Flatest\u002Fdg\u002Fwelcome.html\n- **Security in Amazon Redshift** — https:\u002F\u002Fdocs.aws.amazon.com\u002Fredshift\u002Flatest\u002Fdg\u002Fsecurity.html\n- **Amazon Redshift API Reference** — https:\u002F\u002Fdocs.aws.amazon.com\u002Fredshift\u002Flatest\u002FAPIReference\u002F\n- **AWS CLI Redshift reference** — https:\u002F\u002Fdocs.aws.amazon.com\u002Fcli\u002Flatest\u002Freference\u002Fredshift\u002F\n- **Amazon Redshift Management Guide** — https:\u002F\u002Fdocs.aws.amazon.com\u002Fredshift\u002Flatest\u002Fmgmt\u002Fwelcome.html\n- **Redshift Serverless** — https:\u002F\u002Fdocs.aws.amazon.com\u002Fredshift\u002Flatest\u002Fmgmt\u002Fserverless-whatis.html\n","# Redshift Cluster Auditor\n\n## Mindset\n\n**One-line takeaway:** the verdict is always the **worst** finding across\nsix ordered dimensions, and `PubliclyAccessible: true` on a Redshift\ncluster is in a class of its own — it places a petabyte-scale data\nwarehouse on the public internet with no VPC boundary in front of it.\n\nRedshift clusters concentrate an organisation's most sensitive analytics\ndata: customer records, financial transactions, product analytics. Three\nRedshift-specific behaviours drive the classification:\n\n- **`PubliclyAccessible: true` is a direct internet endpoint.** Unlike RDS\n  (where the SG + PubliclyAccessible + subnet routing combine), Redshift\n  with `PubliclyAccessible: true` is reachable from `0.0.0.0\u002F0` over TCP\n  5439 (or 5440). There is no \"public subnet routing\" subtlety — the flag\n  is the exposure. A public Redshift cluster is a public data warehouse.\n- **Encryption-at-rest is immutable per cluster.** Unlike RDS or EBS,\n  there is no online modify flow and no snapshot-restore toggle. The\n  *only* path is: provision a new encrypted cluster, `UNLOAD` from the\n  old, `COPY` into the new, redirect downstream BI tools, decommission\n  the old. Treat `Encrypted: false` as a migration backlog, not a CLI\n  toggle.\n- **`require_ssl` lives in the parameter group, not the cluster metadata.**\n  The default value is `false` and the default parameter group\n  (`default.redshift-1.0`) cannot be modified — a cluster on the default\n  PG is reporting the engine default, not an operator decision. The\n  parameter must be set to `true` in a *custom* parameter group, and the\n  cluster must be rebooted for the change to apply.\n\n## Quick reference — verdict priority\n\n| Condition | Verdict | Step |\n|---|---|---|\n| `PubliclyAccessible: true` | **PUBLIC** | Step 1 |\n| `Encrypted: false` | **NO_ENCRYPTION** | Step 2 |\n| `require_ssl` parameter value is `false` (or unset) | **NO_SSL** | Step 3 |\n| `LoggingEnabled: false` (describe-logging-status) | **NO_AUDIT_LOG** | Step 4 |\n| Any of: snapshot retention 0, enhanced VPC routing off, SG ingress `0.0.0.0\u002F0` on cluster port | **CONFIG_GAP** | Step 5 |\n| All dimensions pass | **OK** | Step 6 |\n\nApply in order; the first matching verdict wins. The ordering is\ndeliberate: PUBLIC > NO_ENCRYPTION > NO_SSL > NO_AUDIT_LOG > CONFIG_GAP > OK.\nAll findings still appear in the FINDINGS list regardless of which one\nbecame the verdict. See the ordered steps below for Redshift-specific edge\ncases (AWS-managed vs customer-managed KMS, default vs custom parameter\ngroup, Serverless deferral, HSM legacy clusters).\n\n## Pre-flight: cluster metadata gate (run before classification)\n\nBefore evaluating any dimension, classify the cluster itself. Several\ncluster attributes **short-circuit** the audit — misclassifying them\nproduces false positives that erode operator trust.\n\n**Live-account pre-flight checks (skip if doing offline metadata audit):**\n\n1. **Confirm the cluster exists and is reachable.** `aws redshift\n   describe-clusters --cluster-identifier \u003Cid> --profile \u003Cp>` — fail\n   closed (skip the audit, surface an ERROR) if it returns\n   `ClusterNotFound`. A cluster in `deleting` or `final-snapshot` state\n   cannot be remediated in place; note the lifecycle state in the\n   verdict.\n2. **Verify the caller can run `redshift:Describe*` AND\n   `ec2:DescribeSecurityGroups`.** The SG dimension requires\n   cross-service access — a read-only auditor role scoped to Redshift\n   only will silently emit CONFIG_GAP false positives because it cannot\n   see the SG rules. Verify BEFORE the audit, not after.\n3. **Snapshot the cluster's parameter group application.** Reboot\n   windows are required for parameter-group changes that affect SSL.\n   Confirm whether a recent reboot is in flight\n   (`ClusterParameterGroupStatus: Applying` means pending reboot).\n4. **Confirm whether audit logging was ever enabled then disabled.**\n   `describe-logging-status` returns `LoggingEnabled: false` for both\n   \"never enabled\" and \"previously enabled, now off\". For forensics,\n   check the bucket for prior log objects under the cluster's prefix.\n\n| Attribute | Value | Effect on audit |\n|---|---|---|\n| `ClusterStatus` | `available`, `modifying`, `rebooting` | Proceed normally. `modifying`\u002F`rebooting` may mean a parameter-group change is mid-apply. |\n| `ClusterStatus` | `deleting`, `final-snapshot` | Cluster is being decommissioned — emit verdict with a note that remediation is N\u002FA until the cluster is recreated. |\n| `ClusterStatus` | `paused` (RA3 limit, rare) | Cluster is administratively paused. Encryption \u002F SSL \u002F logging dimensions still audit. |\n| `NumberOfNodes: 1` | Single-node | Single-node clusters do **not** provide continuous availability during node failure. Note as an operational risk (not a verdict driver). Multi-AZ deployment (the `MultiAZ` flag on supported node types) is the availability dimension — separate from security. |\n| `NodeType: serverless` or input is a Redshift Serverless workgroup | **Skip this skill.** Redshift Serverless has a different API surface (`redshift-serverless get-workgroup`, no `PubliclyAccessible`, no `ClusterParameterGroupName` — SSL is `require_ssl` on the config). Defer to a Serverless-specific audit. Emit ERROR with a routing note. |\n| `HsmClientCertificateIdentifier` present | **Legacy HSM-encrypted cluster.** Pre-2017 Redshift supported on-prem HSM encryption. HSM-managed clusters report `Encrypted: true` but bypass KMS entirely. Treat as encrypted for verdict purposes; flag as a deprecation risk (HSM integration is end-of-life). |\n| `ElasticIpStatus: ElasticIp` present | **EIP attached.** A public IP pinned to the cluster. Combined with `PubliclyAccessible: true`, this is a stable, DNS-portable public database endpoint — worse than an ephemeral public IP because it survives stop\u002Fstart. |\n| `KmsKeyId: null` AND `Encrypted: true` | **AWS-managed KMS key** (`aws\u002Fredshift`). Encryption is on; you do NOT control rotation or cross-account policy. Note in findings — CMK migration is a hardening step, not a security gap. |\n| `KmsKeyId: \u003Carn>` AND `Encrypted: true` | **Customer-managed KMS key.** Preferred posture. Cross-reference the key policy via the kms-key-policy-auditor skill. |\n| `ClusterParameterGroupName: default.redshift-1.0` (or `.0` for any engine version) | **Default parameter group.** Cannot be modified — `require_ssl` and `enable_user_activity_logging` cannot be set to `true` on the default PG. Any non-default parameter value requires migrating to a custom PG. |\n\n**If the cluster metadata is malformed** (missing `ClusterIdentifier`,\nmissing `PubliclyAccessible`, missing `Encrypted`), output:\n\n```text\nCLUSTER: \u003Cid-or-unknown>\nVERDICT: ERROR\nREASON: Cluster metadata is incomplete — cannot classify (missing \u003Cfield>).\nREMEDIATION: Re-fetch with `aws redshift describe-clusters --cluster-identifier \u003Cid> --output json` and re-audit.\n```\n\n## Process — Classification logic (apply in order, first match wins)\n\n### Step 0: Expert knowledge — non-obvious Redshift behaviours that change classification\n\nFourteen non-obvious Redshift behaviours (PubliclyAccessible vs EnhancedVPCRouting, require_ssl scope, read-only default parameter group, AWS-managed key, encryption immutability, snapshot retention expiry and range, enable-logging vs enable_user_activity_logging, port 5439, cross-region snapshot copy, paused clusters, single-node exposure, CloudTrail control-plane only): moved verbatim to [references\u002Fadvanced-patterns.md](references\u002Fadvanced-patterns.md).\nLoad before classification; Steps 1-6 below assume these constraints.\n\n### Step 1: Public accessibility (highest priority — internet-exposed data warehouse)\n\nIf `PubliclyAccessible: true`:\n\n- **Verdict: PUBLIC.** The cluster is assigned a public IP and accepts\n  connections from `0.0.0.0\u002F0` on the cluster port (subject to SG\n  rules). Even if the SG restricts ingress to a corporate CIDR, the\n  cluster responds to public TCP probes (SYN-ACK) — it is discoverable\n  by any internet scanner, and any future SG widening instantly exposes\n  it. A public Redshift cluster is a public petabyte-scale database.\n- The presence of an Elastic IP (`ElasticIpStatus.ElasticIp` set) makes\n  the exposure *stable* — the public IP survives stop\u002Fstart. Note this\n  in FINDINGS; it does not change the verdict but raises the\n  operational risk.\n\nPubliclyAccessible is evaluated first because it is the only dimension\nwhere the cluster is **directly attackable from anywhere on the\ninternet**. Encryption, SSL, and logging are all important but\npresuppose the cluster is reachable; PUBLIC short-circuits that\nassumption.\n\n### Step 2: Encryption-at-rest (immutable post-creation)\n\nIf `Encrypted: false`:\n\n- **Verdict: NO_ENCRYPTION.** Data-at-rest is unencrypted. Snapshots\n  inherit the cluster's encryption state (an unencrypted cluster\n  produces unencrypted snapshots; you can copy a snapshot with\n  encryption enabled, which is one path to a migration). The KMS key\n  flag is irrelevant when `Encrypted: false`.\n\nIf `Encrypted: true`:\n- With `KmsKeyId: null` → AWS-managed key (`aws\u002Fredshift`). OK for this\n  dimension. Note as an additive finding if CMK is required by the\n  compliance framework.\n- With `KmsKeyId: \u003Carn>` → customer-managed key. OK for this dimension.\n  Recommend cross-referencing the key policy via the\n  kms-key-policy-auditor skill (the key's policy is a separate audit\n  surface).\n\nEncryption is evaluated second because it is the second-highest blast\nradius (after internet exposure): if an unencrypted snapshot leaks or\nthe underlying storage is improperly decommissioned, the data is\nplaintext.\n\n### Step 3: require_ssl parameter group enforcement\n\nInspect the parameter group attached to the cluster\n(`ClusterParameterGroupName`) and its parameter list. Find the\n`require_ssl` parameter.\n\n- **`require_ssl: false` (any source)** → **Verdict: NO_SSL.** SQL\n  clients MAY connect over plaintext. Compliance frameworks (PCI-DSS\n  4.1, HIPAA Transport Encryption, SOC 2 CC6.1) require TLS in transit.\n  Redshift's wire protocol supports SSL; `require_ssl: true` enforces\n  it.\n\n- **`require_ssl: true`** → OK for this dimension.\n- **`require_ssl` parameter absent from the response** → treat as\n  `false` (the engine default is false). Flag as NO_SSL.\n\n**Default parameter group trap:** if the cluster is on\n`default.redshift-1.0` and `Source: engine-default`, the operator has\nnot actively decided. Remediation requires migrating to a custom PG —\nemphasise this in REMEDIATION. A custom PG with `require_ssl: false`\nand `Source: user` is an explicit operator decision to allow plaintext\n(a deliberate misconfiguration, not a default).\n\n**Pending-reboot trap:** if `ClusterParameterGroupStatus: Applying` or\n`pending-reboot`, a recent parameter change has not yet taken effect.\nThe current effective value is the OLD value, not the new one. Surface\nthe pending value in the FINDINGS; the verdict reflects the current\neffective state.\n\n### Step 4: Audit logging (S3 export)\n\nInspect `describe-logging-status`. The key field is `LoggingEnabled`.\n\n- **`LoggingEnabled: false`** → **Verdict: NO_AUDIT_LOG.** Connection,\n  user-activity, and DDL events are not being exported to S3. This\n  eliminates forensic capability for \"who ran what query when\" and\n  breaks compliance evidence chains (PCI-DSS 10.2, HIPAA §164.312(b),\n  SOC 2 CC7.2).\n\n- **`LoggingEnabled: true`** → OK for this dimension. Note the bucket\n  name and prefix in FINDINGS; recommend verifying the bucket's\n  object-lock \u002F retention policy and access controls separately.\n\nAudit logging is evaluated fourth because it is the forensic\nprerequisite for investigating the other findings. Without it, a\nPUBLIC+NO_ENCRYPTION breach cannot be reconstructed.\n\n### Step 5: Configuration gaps (additive findings)\n\nEvaluate the remaining dimensions and surface any failures as CONFIG_GAP\nfindings. If any of the following fail AND Steps 1-4 all passed, the\nverdict is **CONFIG_GAP**.\n\n| Sub-check | Failure condition | Why it matters |\n|---|---|---|\n| **Automated snapshots** | `AutomatedSnapshotRetentionPeriod: 0` | PITR disabled. Existing automated snapshots expire within the next snapshot window. Manual snapshots are unaffected but require active management. |\n| **Enhanced VPC routing** | `EnhancedVPCRouting: false` | `COPY` \u002F `UNLOAD` traffic bypasses VPC SGs, NACLs, and VPC endpoints — data path is not subject to network controls. Engine default is `false`. |\n| **VPC security group ingress on cluster port** | Any SG rule with `0.0.0.0\u002F0` (or `::\u002F0`) on the cluster's TCP port (default 5439, configurable via `Endpoint.Port`) | Even with `PubliclyAccessible: false`, an `0.0.0.0\u002F0` SG rule on the cluster port is a CONFIG_GAP — a future flip of `PubliclyAccessible` to `true` instantly exposes the cluster. Note: SG ingress does NOT make a private cluster public; it expands the in-VPC attack surface. |\n| **User-activity logging** | `enable_user_activity_logging: false` parameter value | STL_QUERY retains query text for ~3-7 days; without user_activity_logging exported to S3 (via the audit log pipeline), long-horizon forensic queries are impossible. Treat as CONFIG_GAP (additive to Step 4 — if Step 4 failed, this is folded into NO_AUDIT_LOG). |\n| **Snapshot retention below compliance floor** | `AutomatedSnapshotRetentionPeriod: 1` on a production cluster | The engine default is 1 day. PCI\u002FHIPAA typically require 7+; flag as a soft CONFIG_GAP if compliance framework applies. |\n| **Cross-region snapshot copy absent** | `ClusterSnapshotCopyStatus` is absent on a production cluster | Single-region snapshots = no DR. Flag as a resilience CONFIG_GAP (not security). |\n\nFor the SG check specifically: the SG dimension surfaces the *exposure*\nof the cluster within its VPC. SG rules from corporate CIDRs\n(`10.0.0.0\u002F8`, `172.16.0.0\u002F12`, `192.168.0.0\u002F16`) are typically fine;\n`0.0.0.0\u002F0` on the cluster port is the canonical misconfiguration.\n\n### Step 6: Aggregation — OK verdict\n\nIf all dimensions pass (PUBLIC not set, Encrypted true, require_ssl true,\nLoggingEnabled true, no CONFIG_GAP sub-findings), the verdict is **OK**.\n\n```text\nverdict_priority = PUBLIC > NO_ENCRYPTION > NO_SSL > NO_AUDIT_LOG > CONFIG_GAP > OK\n```\n\nApply in order; the first matching step wins. ALL findings (including\nnon-verdict-driving ones) appear in the FINDINGS list.\n\n## Output format (per cluster)\n\n```text\nCLUSTER: \u003Ccluster-identifier>\nVERDICT: PUBLIC | NO_ENCRYPTION | NO_SSL | NO_AUDIT_LOG | CONFIG_GAP | OK\nREASON: \u003C1-2 sentences citing the worst finding and its step number>\nFINDINGS:\n  - [PUBLIC] \u003Cfinding description (Step 1)>\n  - [NO_SSL] \u003Cfinding description (Step 3)>\n  - [OK] \u003Cdimension that passed>\nREMEDIATION: \u003Cspecific action per finding, or \"None required\" if OK>\n```\n\n### Worked example — public cluster with encryption disabled\n\n```text\nCLUSTER: public-cluster-no-encryption\nVERDICT: PUBLIC\nREASON: PubliclyAccessible is true, placing the cluster on the public\ninternet on TCP 5439 (Step 1). Encryption is also disabled, require_ssl\nis false, and S3 audit logging is off — every dimension fails.\nFINDINGS:\n  - [PUBLIC] PubliclyAccessible: true — cluster accepts connections from\n    0.0.0.0\u002F0 on the cluster port (Step 1)\n  - [NO_ENCRYPTION] Encrypted: false — data-at-rest is plaintext (Step 2)\n  - [NO_SSL] require_ssl: false in default.redshift-1.0 — clients may\n    connect over plaintext (Step 3)\n  - [NO_AUDIT_LOG] LoggingEnabled: false — no S3 audit trail (Step 4)\n  - [CONFIG_GAP] EnhancedVPCRouting: false — COPY\u002FUNLOAD bypasses VPC\n    controls (Step 5)\n  - [CONFIG_GAP] SG sg-0aaa11112222 has 0.0.0.0\u002F0 on TCP 5439 (Step 5)\nREMEDIATION:\n  1. PUBLIC — Set PubliclyAccessible to false: see CLI below. This\n     requires a client reboot. If public access is INTENTIONAL (e.g., a\n     partner-data workload), restrict the SG to specific partner CIDRs\n     immediately.\n  2. NO_ENCRYPTION — Plan an encryption migration: provision a new\n     encrypted cluster, UNLOAD to S3, COPY into the new cluster, redirect\n     downstream tools, decommission the old cluster. Encryption cannot be\n     toggled on an existing cluster.\n  3. NO_SSL — Create a custom parameter group with require_ssl: true,\n     associate it with the cluster, and reboot for the change to take\n     effect.\n  4. NO_AUDIT_LOG — Enable S3 audit logging:\n     aws redshift enable-logging --cluster-identifier public-cluster-no-encryption \\\n       --log-destination-name S3 --bucket-name audit-logs-111111111111 \\\n       --s3-key-prefix redshift\u002Fpublic-cluster-no-encryption\u002F\n  5. CONFIG_GAP — Enable EnhancedVPCRouting (requires cluster reboot) and\n     restrict the SG ingress to known CIDRs on the cluster port.\n```\n\n## Edge-case handling\n\nNine edge cases (modifying\u002Frebooting states, default parameter group, Multi-AZ, RA3\u002FDC2\u002FDS2 node types, HSM legacy, snapshot copy status, Serverless deferral, AWS-managed key, non-default port): moved verbatim to [references\u002Fadvanced-patterns.md](references\u002Fadvanced-patterns.md).\nConsult after the ordered steps when the cluster metadata is unusual.\n\n## Anti-Patterns — NEVER\n\n- NEVER classify `PubliclyAccessible: true` as anything other than\n  PUBLIC. The flag is a direct internet endpoint on TCP 5439\u002F5440. SG\n  rules may restrict ingress, but the cluster is still discoverable and\n  attackable by any internet scanner. A public data warehouse is the\n  highest-impact Redshift misconfiguration.\n\n- NEVER recommend \"just enable encryption\" on an existing cluster.\n  Redshift encryption is immutable per cluster. The only remediation\n  is a new cluster + UNLOAD\u002FCOPY migration. Stating otherwise causes\n  operators to attempt `modify-cluster --encrypted` (which does not\n  exist) and lose trust in the audit.\n\n- NEVER treat `require_ssl: false` from the default parameter group as\n  an operator decision. `Source: engine-default` means the operator\n  has not hardened this — they may not even know the parameter exists.\n  Remediation requires migrating to a custom PG (the default PG is\n  read-only).\n\n- NEVER assume `enable-logging` and `enable_user_activity_logging` are\n  the same thing. `enable-logging` (the API) exports audit events to\n  S3. `enable_user_activity_logging` (the parameter) logs every SQL\n  query to STL_QUERY. They are separate controls with separate\n  forensic value. Treat absence of either as a gap.\n\n- NEVER assume CloudTrail covers Redshift SQL activity. CloudTrail logs\n  control-plane events (CreateCluster, ModifyCluster, DeleteCluster).\n  It does NOT log data-plane SQL queries — that is exclusively the\n  domain of `enable-logging` S3 audit logs. \"We have CloudTrail\" does\n  not satisfy \"we audit database activity\" for Redshift.\n\n- NEVER conflate `PubliclyAccessible: false` with \"no internet exposure\n  for data transfers\". `EnhancedVPCRouting: false` (the default) means\n  COPY\u002FUNLOAD traffic to S3 traverses the public AWS network path,\n  bypassing your VPC's SGs, NACLs, and VPC endpoints. Privacy of the\n  cluster endpoint and privacy of the bulk-transfer data path are\n  independent concerns.\n\n- NEVER treat a single-node cluster as having high availability.\n  Single-node clusters have no replication and no failover. Destructive\n  queries (DROP TABLE, TRUNCATE) affect the entire dataset with no\n  redundant copy. Note this as an operational risk; do not let it\n  drive the security verdict.\n\n- NEVER flag `KmsKeyId: null` with `Encrypted: true` as an encryption\n  gap. This is the AWS-managed key (`aws\u002Fredshift`) — encryption is\n  on. Rotates annually on AWS's schedule. Acceptable for most\n  compliance frameworks. Flag CMK as a hardening opportunity, not a\n  finding.\n\n- NEVER attempt to apply this skill to a Redshift Serverless workgroup.\n  Serverless has a different API surface (no `PubliclyAccessible`, no\n  `ClusterParameterGroupName`, SSL enforced via `config-parameters` on\n  the workgroup). Emit ERROR with a routing note — misclassifying a\n  Serverless workgroup as PUBLIC or NO_SSL is a false positive.\n\n- NEVER assume `AutomatedSnapshotRetentionPeriod: 0` only disables\n  future snapshots. Redshift immediately begins expiring existing\n  automated snapshots. Operators setting retention to 0 to \"save cost\"\n  without first converting critical snapshots to manual lose PITR\n  irreversibly within hours.\n\n- NEVER recommend modifying a default parameter group. The default PG\n  (`default.redshift-1.0`) is read-only — `modify-cluster-parameter-group`\n  fails. Remediation requires creating a custom PG and associating it\n  with the cluster.\n\n- NEVER treat an SG with `0.0.0.0\u002F0` on TCP 5439 as making the cluster\n  public. The cluster is public ONLY if `PubliclyAccessible: true`. SG\n  `0.0.0.0\u002F0` ingress on a private cluster is a CONFIG_GAP — it\n  expands the in-VPC attack surface and sets up a future exposure if\n  the flag is flipped, but it does not by itself expose the cluster to\n  the internet.\n\n- NEVER evaluate the SG ingress on the wrong port. The cluster port is\n  in `Endpoint.Port` (default 5439, customisable up to 65535). SG\n  rules matching TCP 5432 (Postgres default) silently fail to permit\n  traffic. Read the actual port from the metadata, not the default.\n\n- NEVER flag a cluster in `paused` state as safer or out-of-scope.\n  Paused clusters retain their encryption, SG, and logging posture;\n  the audit is identical. Pause is a cost-optimisation, not a security\n  control.\n\n## Pre-flight safety checks (run before any remediation CLI)\n\nSeven pre-flight safety checks (CONFIRM gate, reboot warnings, encryption migration scope, pre-change snapshot, audit bucket ownership, cross-region cost, shared-SG verification): moved verbatim to [references\u002Fdiagnostic-commands.md](references\u002Fdiagnostic-commands.md).\nDefense-in-depth before any remediation CLI.\n\n## Remediation guidance\n\n**Remediation ordering principle:** additive changes first (enable\nlogging, associate a new PG), destructive \u002F state-changing changes last\n(modify cluster, reboot). Public accessibility and encryption are the\ntwo findings that cannot be silently fixed in place — plan migrations,\nnot toggles.\n\nPer-verdict remediation playbooks with full CLI blocks (PUBLIC, NO_ENCRYPTION UNLOAD\u002FCOPY migration, NO_SSL custom parameter group, NO_AUDIT_LOG enable-logging, CONFIG_GAP sub-findings, OK follow-ups): moved verbatim to [references\u002Fremediation-procedures.md](references\u002Fremediation-procedures.md).\nEmit per-finding remediation from that reference.\n\n## Recent AWS features (2024-2026)\n\nFour recent AWS features 2024-2026 (Serverless enhancements, Aurora Zero-ETL, data sharing, Redshift ML): moved verbatim to [references\u002Fadvanced-patterns.md](references\u002Fadvanced-patterns.md).\nConsult before citing feature limits or recency.\n\n## Related skills\n\n- **kms-key-policy-auditor:** audit the KMS key policy when a\n  customer-managed key encrypts the cluster. The cluster audit treats\n  the key as a black box; the key-policy audit determines who can\n  decrypt cluster snapshots.\n- **ec2-security-group-auditor:** deeper SG analysis including\n  prefix-list composition and cross-VPC exposure (the cluster audit\n  checks only the cluster port's exposure on the attached SGs).\n- **s3-public-access-auditor:** audit the S3 bucket receiving UNLOAD\n  exports or audit logs — a private cluster with a public UNLOAD bucket\n  silently exfiltrates data.\n\n## References (load on demand)\n\n- [references\u002Fadvanced-patterns.md](references\u002Fadvanced-patterns.md) — Step 0 non-obvious behaviours, edge-case handling, recent AWS features (2024-2026).\n- [references\u002Fdiagnostic-commands.md](references\u002Fdiagnostic-commands.md) — Pre-flight safety checks before any remediation CLI.\n- [references\u002Fremediation-procedures.md](references\u002Fremediation-procedures.md) — Per-verdict remediation playbooks with CLI blocks.\n\n## Domain\n\nAWS CloudOps \u002F Redshift Analytics Security & Compliance.\n\n## AWS documentation\n\n- **Amazon Redshift documentation** — https:\u002F\u002Fdocs.aws.amazon.com\u002Fredshift\u002Flatest\u002Fdg\u002Fwelcome.html\n- **Security in Amazon Redshift** — https:\u002F\u002Fdocs.aws.amazon.com\u002Fredshift\u002Flatest\u002Fdg\u002Fsecurity.html\n- **Amazon Redshift API Reference** — https:\u002F\u002Fdocs.aws.amazon.com\u002Fredshift\u002Flatest\u002FAPIReference\u002F\n- **AWS CLI Redshift reference** — https:\u002F\u002Fdocs.aws.amazon.com\u002Fcli\u002Flatest\u002Freference\u002Fredshift\u002F\n- **Amazon Redshift Management Guide** — https:\u002F\u002Fdocs.aws.amazon.com\u002Fredshift\u002Flatest\u002Fmgmt\u002Fwelcome.html\n- **Redshift Serverless** — https:\u002F\u002Fdocs.aws.amazon.com\u002Fredshift\u002Flatest\u002Fmgmt\u002Fserverless-whatis.html\n",{"name":5,"license":15,"metadata":24,"description":7,"compatibility":39},{"tags":25,"phase":26,"author":27,"domain":28,"family":29,"version":30,"keywords":31,"complexity":32,"entry_point":33,"when_to_use":34,"requires_llm":35,"verdict_shape":36,"invocation_schema":37,"supports_pipeline":35,"activation_triggers":38},"redshift, analytics, security, encryption, ssl, audit-logging, snapshots, vpc, compliance, audit","2","Jacky Chan — AWS Community Builder","aws-cloudops","Analytics","0.1.0","Redshift, data warehouse, PubliclyAccessible, encryption, KMS, require_ssl, SSL, TLS, parameter group, audit logging, S3 audit log, enable_user_activity_logging, automated snapshots, AutomatedSnapshotRetentionPeriod, enhanced VPC routing, EnhancedVPCRouting, COPY, UNLOAD, VPC security group, cluster port 5439, data-warehouse hardening, compliance, Redshift Serverless","high","false","Reviewing a Redshift provisioned cluster before production deployment, auditing encryption-at-rest, validating require_ssl enforcement, checking S3 audit logging coverage, inspecting automated-snapshot retention, evaluating enhanced VPC routing posture, or hardening a data warehouse for a compliance review.","true","PUBLIC | NO_ENCRYPTION | NO_SSL | NO_AUDIT_LOG | CONFIG_GAP | OK","Input: either (a) a describe-clusters Cluster block, optionally paired with describe-logging-status, describe-cluster-parameters (for the attached parameter group), and describe-security-groups outputs, OR (b) a ClusterIdentifier for live-account audit. Output: deterministic CLUSTER\u002FVERDICT\u002FREASON\u002FFINDINGS\u002FREMEDIATION block per cluster, where VERDICT ∈ {PUBLIC, NO_ENCRYPTION, NO_SSL, NO_AUDIT_LOG, CONFIG_GAP, OK, ERROR}.","audit this Redshift cluster, is my Redshift cluster public, is encryption enabled on Redshift, is require_ssl on, is Redshift audit logging configured, are automated snapshots enabled, enhanced VPC routing Redshift, Redshift security group audit, harden this data warehouse","Agent runtime that reads SKILL.md (Claude Code, Cursor, Windsurf, Codex, Gemini). No AWS CLI required for offline metadata classification. Live-account audits use aws redshift describe-clusters, describe-logging-status, describe-cluster-parameter-groups (with describe-cluster-parameters for the PG), aws ec2 describe-security-groups, and aws kms describe-key (AWS CLI v2, SSO or key-based credentials).","main","eeace51bd5dd1db2030729d83eeab988861553aa","skills\u002Fredshift-cluster-auditor",0,"https:\u002F\u002Fgithub.com\u002Fjhkchan\u002Faws-agent-skills","https:\u002F\u002Fgithub.com\u002Fjhkchan","CTO & Co-Founder @Beever-AI  & @votee. CAISP. AWS Community Builder. Organizer@GDGCloudHK. MongoDB Champion. Technopreneur + Infornographer.","2026-09-24 08:54:10",{"source":8,"skillName":5,"commands":49,"prompt":59,"zipAvailable":60},[50,53,56],{"tool":51,"command":52},"npx","npx skills add jhkchan\u002Faws-agent-skills --skill redshift-cluster-auditor",{"tool":54,"command":55},"bunx","bunx skills add jhkchan\u002Faws-agent-skills --skill redshift-cluster-auditor",{"tool":57,"command":58},"pnpm","pnpm dlx skills add jhkchan\u002Faws-agent-skills --skill redshift-cluster-auditor","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Fjhkchan\u002Faws-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Fredshift-cluster-auditor\u002FSKILL.md",true,[62,68,72,76,80,84,88,92,96,100,104,108,112,116,120,124,128,132,136,140],{"path":63,"resourceType":64,"resourceTypeDesc":65,"sizeBytes":66,"sourceHtmlUrl":67},"eval\u002Ftest-cases.yaml","REFERENCE","参考文档",12820,"https:\u002F\u002Fgithub.com\u002Fjhkchan\u002Faws-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Fredshift-cluster-auditor\u002Feval\u002Ftest-cases.yaml",{"path":69,"resourceType":64,"resourceTypeDesc":65,"sizeBytes":70,"sourceHtmlUrl":71},"evals\u002Fbaselines\u002Faudit-logging-disabled-clean.md",983,"https:\u002F\u002Fgithub.com\u002Fjhkchan\u002Faws-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Fredshift-cluster-auditor\u002Fevals\u002Fbaselines\u002Faudit-logging-disabled-clean.md",{"path":73,"resourceType":64,"resourceTypeDesc":65,"sizeBytes":74,"sourceHtmlUrl":75},"evals\u002Fbaselines\u002Fconfig-gap-no-snapshots-no-evr.md",1060,"https:\u002F\u002Fgithub.com\u002Fjhkchan\u002Faws-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Fredshift-cluster-auditor\u002Fevals\u002Fbaselines\u002Fconfig-gap-no-snapshots-no-evr.md",{"path":77,"resourceType":64,"resourceTypeDesc":65,"sizeBytes":78,"sourceHtmlUrl":79},"evals\u002Fbaselines\u002Ffully-hardened-ok.md",1052,"https:\u002F\u002Fgithub.com\u002Fjhkchan\u002Faws-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Fredshift-cluster-auditor\u002Fevals\u002Fbaselines\u002Ffully-hardened-ok.md",{"path":81,"resourceType":64,"resourceTypeDesc":65,"sizeBytes":82,"sourceHtmlUrl":83},"evals\u002Fbaselines\u002Fno-ssl-default-param-group.md",959,"https:\u002F\u002Fgithub.com\u002Fjhkchan\u002Faws-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Fredshift-cluster-auditor\u002Fevals\u002Fbaselines\u002Fno-ssl-default-param-group.md",{"path":85,"resourceType":64,"resourceTypeDesc":65,"sizeBytes":86,"sourceHtmlUrl":87},"evals\u002Fbaselines\u002Fprivate-unencrypted.md",931,"https:\u002F\u002Fgithub.com\u002Fjhkchan\u002Faws-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Fredshift-cluster-auditor\u002Fevals\u002Fbaselines\u002Fprivate-unencrypted.md",{"path":89,"resourceType":64,"resourceTypeDesc":65,"sizeBytes":90,"sourceHtmlUrl":91},"evals\u002Fbaselines\u002Fpublic-cluster-no-encryption.md",1192,"https:\u002F\u002Fgithub.com\u002Fjhkchan\u002Faws-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Fredshift-cluster-auditor\u002Fevals\u002Fbaselines\u002Fpublic-cluster-no-encryption.md",{"path":93,"resourceType":64,"resourceTypeDesc":65,"sizeBytes":94,"sourceHtmlUrl":95},"evals\u002Feval_config.yaml",2063,"https:\u002F\u002Fgithub.com\u002Fjhkchan\u002Faws-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Fredshift-cluster-auditor\u002Fevals\u002Feval_config.yaml",{"path":97,"resourceType":64,"resourceTypeDesc":65,"sizeBytes":98,"sourceHtmlUrl":99},"evals\u002Fevals.json",5517,"https:\u002F\u002Fgithub.com\u002Fjhkchan\u002Faws-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Fredshift-cluster-auditor\u002Fevals\u002Fevals.json",{"path":101,"resourceType":64,"resourceTypeDesc":65,"sizeBytes":102,"sourceHtmlUrl":103},"evals\u002Fprompts\u002Faudit-logging-disabled-clean.md",1115,"https:\u002F\u002Fgithub.com\u002Fjhkchan\u002Faws-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Fredshift-cluster-auditor\u002Fevals\u002Fprompts\u002Faudit-logging-disabled-clean.md",{"path":105,"resourceType":64,"resourceTypeDesc":65,"sizeBytes":106,"sourceHtmlUrl":107},"evals\u002Fprompts\u002Fconfig-gap-no-snapshots-no-evr.md",1215,"https:\u002F\u002Fgithub.com\u002Fjhkchan\u002Faws-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Fredshift-cluster-auditor\u002Fevals\u002Fprompts\u002Fconfig-gap-no-snapshots-no-evr.md",{"path":109,"resourceType":64,"resourceTypeDesc":65,"sizeBytes":110,"sourceHtmlUrl":111},"evals\u002Fprompts\u002Ffully-hardened-ok.md",1171,"https:\u002F\u002Fgithub.com\u002Fjhkchan\u002Faws-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Fredshift-cluster-auditor\u002Fevals\u002Fprompts\u002Ffully-hardened-ok.md",{"path":113,"resourceType":64,"resourceTypeDesc":65,"sizeBytes":114,"sourceHtmlUrl":115},"evals\u002Fprompts\u002Fno-ssl-default-param-group.md",1226,"https:\u002F\u002Fgithub.com\u002Fjhkchan\u002Faws-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Fredshift-cluster-auditor\u002Fevals\u002Fprompts\u002Fno-ssl-default-param-group.md",{"path":117,"resourceType":64,"resourceTypeDesc":65,"sizeBytes":118,"sourceHtmlUrl":119},"evals\u002Fprompts\u002Fprivate-unencrypted.md",1131,"https:\u002F\u002Fgithub.com\u002Fjhkchan\u002Faws-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Fredshift-cluster-auditor\u002Fevals\u002Fprompts\u002Fprivate-unencrypted.md",{"path":121,"resourceType":64,"resourceTypeDesc":65,"sizeBytes":122,"sourceHtmlUrl":123},"evals\u002Fprompts\u002Fpublic-cluster-no-encryption.md",1113,"https:\u002F\u002Fgithub.com\u002Fjhkchan\u002Faws-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Fredshift-cluster-auditor\u002Fevals\u002Fprompts\u002Fpublic-cluster-no-encryption.md",{"path":125,"resourceType":64,"resourceTypeDesc":65,"sizeBytes":126,"sourceHtmlUrl":127},"examples\u002Fend-to-end.md",9756,"https:\u002F\u002Fgithub.com\u002Fjhkchan\u002Faws-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Fredshift-cluster-auditor\u002Fexamples\u002Fend-to-end.md",{"path":129,"resourceType":64,"resourceTypeDesc":65,"sizeBytes":130,"sourceHtmlUrl":131},"references\u002Fadvanced-patterns.md",10074,"https:\u002F\u002Fgithub.com\u002Fjhkchan\u002Faws-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Fredshift-cluster-auditor\u002Freferences\u002Fadvanced-patterns.md",{"path":133,"resourceType":64,"resourceTypeDesc":65,"sizeBytes":134,"sourceHtmlUrl":135},"references\u002Fdiagnostic-commands.md",2707,"https:\u002F\u002Fgithub.com\u002Fjhkchan\u002Faws-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Fredshift-cluster-auditor\u002Freferences\u002Fdiagnostic-commands.md",{"path":137,"resourceType":64,"resourceTypeDesc":65,"sizeBytes":138,"sourceHtmlUrl":139},"references\u002Fremediation-procedures.md",5571,"https:\u002F\u002Fgithub.com\u002Fjhkchan\u002Faws-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Fredshift-cluster-auditor\u002Freferences\u002Fremediation-procedures.md",{"path":141,"resourceType":64,"resourceTypeDesc":65,"sizeBytes":142,"sourceHtmlUrl":143},"skill.usf.yaml",4242,"https:\u002F\u002Fgithub.com\u002Fjhkchan\u002Faws-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Fredshift-cluster-auditor\u002Fskill.usf.yaml",[145,152,159,166,173,180],{"skillNo":146,"name":147,"nameZh":6,"description":148,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":149,"categories":150,"keywords":151,"publishedAt":19},"01M7ADVD60X4PCAPKSE58P9NDC","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 \u002F method \u002F 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 \u002F 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 \u002F EKS \u002F 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.","https:\u002F\u002Fgithub.com\u002Fjhkchan\u002Faws-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Fxray-tracing-deployer\u002FSKILL.md",[],[],{"skillNo":153,"name":154,"nameZh":6,"description":155,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":156,"categories":157,"keywords":158,"publishedAt":19},"01M7ADTVJ7DCBM9CB1FY51GH7J","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.","https:\u002F\u002Fgithub.com\u002Fjhkchan\u002Faws-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Fwellarchitected-workload-auditor\u002FSKILL.md",[],[],{"skillNo":160,"name":161,"nameZh":6,"description":162,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":163,"categories":164,"keywords":165,"publishedAt":19},"01M7ADT59G6FTTNZYHT2QRPNNP","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 \u002F 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.","https:\u002F\u002Fgithub.com\u002Fjhkchan\u002Faws-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Fwellarchitected-review-operator\u002FSKILL.md",[],[],{"skillNo":167,"name":168,"nameZh":6,"description":169,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":170,"categories":171,"keywords":172,"publishedAt":19},"01M7ADSH5JA9QHNPPHQCD7EVA7","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\u002FAPI Gateway\u002FCloudFront, or configuring CAPTCHA\u002FChallenge 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.","https:\u002F\u002Fgithub.com\u002Fjhkchan\u002Faws-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Fwafv2-web-acl-deployer\u002FSKILL.md",[],[],{"skillNo":174,"name":175,"nameZh":6,"description":176,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":177,"categories":178,"keywords":179,"publishedAt":19},"01M7ADRYFAT8R73FBNCHRC573X","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\u002Fbypass 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.","https:\u002F\u002Fgithub.com\u002Fjhkchan\u002Faws-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Fwafv2-web-acl-auditor\u002FSKILL.md",[],[],{"skillNo":181,"name":182,"nameZh":6,"description":183,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":184,"categories":185,"keywords":186,"publishedAt":19},"01M7ADR8W7G72JX4KFGN03QK8Q","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.","https:\u002F\u002Fgithub.com\u002Fjhkchan\u002Faws-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Fwaf-rule-deployer\u002FSKILL.md",[],[]]