[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M7ACEX7DGTBR15JASZGFFCWM":3,"skill-same:01M7ACEX7DGTBR15JASZGFFCWM":135},{"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},"01M7ACEX7DGTBR15JASZGFFCWM","service-quotas-usage-auditor",null,"Audits AWS Service Quotas — quota utilization per service, approaching limits (>=80%), CloudWatch alarm coverage on AWS\u002FUsage metrics, applied vs default quota drift, adjustable quotas stuck at default with rising usage, denied or stale quota increase requests, and non-trackable quotas lacking a UsageMetric. Emits a deterministic verdict (APPROACHING_LIMIT | NO_ALARM | CONFIG_GAP | OK) per quota with enumerated findings and CLI remediation. Use when reviewing quota utilization, checking for approaching service limits, auditing quota increase request history, or validating CloudWatch alarm coverage.","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\u002Fservice-quotas-usage-auditor\u002FSKILL.md",[],[],"2026-11-15 14:19:17",false,"---\nname: service-quotas-usage-auditor\ndescription: Audits AWS Service Quotas — quota utilization per service, approaching limits (>=80%), CloudWatch alarm coverage on AWS\u002FUsage metrics, applied vs default quota drift, adjustable quotas stuck at default with rising usage, denied or stale quota increase requests, and non-trackable quotas lacking a UsageMetric. Emits a deterministic verdict (APPROACHING_LIMIT | NO_ALARM | CONFIG_GAP | OK) per quota with enumerated findings and CLI remediation. Use when reviewing quota utilization, checking for approaching service limits, auditing quota increase request history, or validating CloudWatch alarm coverage.\nlicense: Apache-2.0\ncompatibility: Agent runtime that reads SKILL.md (Claude Code, Cursor, Windsurf, Codex, Gemini). No AWS CLI required for offline quota-snapshot classification. Live-account audits use aws service-quotas list-service-quotas, get-service-quota, get-aws-default-service-quota, list-requested-service-quota-change-history, and aws cloudwatch describe-alarbs-for-metric (AWS CLI v2, SSO or key-based credentials).\nmetadata:\n  domain: aws-cloudops\n  complexity: medium\n  requires_llm: 'true'\n  phase: '2'\n  supports_pipeline: 'true'\n  entry_point: 'false'\n  family: Management\n  verdict_shape: APPROACHING_LIMIT | NO_ALARM | CONFIG_GAP | OK\n  when_to_use: Auditing AWS Service Quotas utilization, CloudWatch alarm coverage on AWS\u002FUsage metrics, and quota increase request history.\n  activation_triggers: audit service quotas, check quota utilization, approaching service limit, quota increase request, cloudwatch alarm on quota, service limits audit\n  invocation_schema: 'Input: either (a) a Service Quotas snapshot (quota code, service code, applied value, default value, UsageMetric, utilization, increase request history, CloudWatch alarm state), OR (b) a service-code for live-account audit via aws service-quotas list-service-quotas. Output: deterministic QUOTA\u002FVERDICT\u002FREASON\u002FUTILIZATION\u002FFINDINGS\u002F REMEDIATION block per quota, where VERDICT is one of APPROACHING_LIMIT, NO_ALARM, CONFIG_GAP, OK (or ERROR for malformed input).'\n  version: 0.1.0\n  author: Jacky Chan — AWS Community Builder\n  keywords: Service Quotas, quota utilization, approaching limit, AWS\u002FUsage, CloudWatch alarm, quota increase, applied quota, default quota, Adjustable, UsageMetric, quota code, service limits, Trusted Advisor, quota monitoring, quota remediation\n  tags: service-quotas, management, quota, limits, cloudwatch, monitoring, audit\n---\n\n# Service Quotas Usage Auditor\n\n## Mindset\n\n**One-line takeaway:** the verdict is the **worst** finding across all\ndimensions, and three Service Quotas behaviors are easy to misjudge —\n**applied vs default** (list-service-quotas returns the APPLIED value,\nnot the AWS default), **UsageMetric presence** (quotas without a metric\nblock CANNOT be auto-monitored via CloudWatch), and **request status\nlifecycle** (a PENDING increase has not taken effect — the applied quota\nis still the old value until APPROVED).\n\nA quota is the hard ceiling on a specific AWS resource or API rate.\nExhausting one is a **silent outage**: API calls return\n`ThrottlingException` or `LimitExceededException` with no remediation\nwindow. Unlike a misconfigured policy (fixable by editing text), a quota\nthat blocks operations can stall an entire deployment pipeline while a\nsupport case grinds through the approval process.\n\n## Quick reference — verdict thresholds\n\n| Condition | Verdict | Step |\n|---|---|---|\n| Utilization >= 80% of applied quota (inclusive: exactly 80% counts) | **APPROACHING_LIMIT** | 1 |\n| `Adjustable: false` AND utilization >= 50% | **CONFIG_GAP** | 2b |\n| Increase request DENIED\u002FNOT_APPROVED AND utilization >= 50% | **CONFIG_GAP** | 2c |\n| At AWS default, `Adjustable: true`, utilization >= 50%, request history completely empty | **CONFIG_GAP** | 2d |\n| No `UsageMetric` block defined | **CONFIG_GAP** | 2a |\n| Has `UsageMetric`, utilization \u003C 80%, no CloudWatch alarm | **NO_ALARM** | 3 |\n| Utilization \u003C 50%, alarm in place or N\u002FA | **OK** | 4 |\n\nSee the ordered steps below for edge cases. Deep Service Quotas +\nCloudWatch integration details are in the\n[Deep reference](#deep-reference-service-quotas--cloudwatch-internals)\nsection at the end.\n\n## Pre-flight: quota metadata gate (run before classification)\n\nBefore evaluating utilization, classify the quota metadata itself.\nSeveral attributes **short-circuit** the audit — misclassifying them\nproduces false positives.\n\n| Attribute | Value | Effect on audit |\n|---|---|---|\n| `UsageMetric` | absent | **No CloudWatch metric exists.** Cannot auto-monitor utilization. This is a CONFIG_GAP (Step 2a) regardless of current utilization — you are flying blind. |\n| `UsageMetric` | present | Proceed — utilization is auto-trackable via `AWS\u002FUsage`. |\n| `Adjustable` | `false` | **Hard limit.** Cannot request an increase via `request-service-quota-increase`. If utilization >= 50%, the only remediation is architectural (distribute load, use a different service, reduce usage). Flag as CONFIG_GAP (Step 2b). |\n| `Adjustable` | `true` | Quota is increasable. Proceed to check increase request history. |\n| `GlobalQuota` | `true` | **Account-wide quota** (e.g., IAM users, S3 buckets). Applies across all regions. An increase applies globally. |\n| `GlobalQuota` | `false` | **Regional quota** (e.g., VPCs per region, EC2 vCPUs per region). An increase in us-east-1 does NOT apply to eu-west-1 — request per-region. |\n| `Unit` | `Gigabytes`, `Mbps`, etc. | **Unit mismatch risk.** Never compare a utilization value in one unit against a quota in another. A quota of \"100 Gigabytes\" at \"50 Count\" usage is a data error. |\n| Applied == Default | — | Quota has never been increased. Not inherently bad, but combined with utilization >= 50% and `Adjustable: true`, it signals that no proactive capacity management has occurred (Step 2d). |\n| Applied > Default | — | An increase was APPROVED at some point — active capacity management. This is a positive signal. |\n\n**Multi-service \u002F account-wide sweep note (pagination):** when auditing\nevery quota for a service, `aws service-quotas list-service-quotas\n--service-code \u003Ccode>` returns at most 100 per page. Use\n`--starting-token` from the prior `NextToken` to page through all\nquotas. The same applies to\n`list-requested-service-quota-change-history`. Always drain the\npaginator to completion — the long tail often contains the quota that\nmatters.\n\n**Malformed input handling.** If the quota snapshot is missing the\napplied value, missing quota code, has a utilization value not\nparseable as a number, or has a unit mismatch between utilization and\nquota, output an ERROR block and refuse classification. Do NOT guess\nthe utilization percentage — a wrong percentage produces a wrong\nverdict. Output:\n\n```text\nQUOTA: \u003Cquota-code or name>\nVERDICT: ERROR\nREASON: Quota snapshot is incomplete or malformed — cannot classify.\nREMEDIATION: Retrieve the canonical data with `aws service-quotas\nget-service-quota --service-code \u003Ccode> --quota-code \u003Ccode> --output json` and re-audit.\n```\n\n## Process — Classification logic (apply in order, aggregate worst)\n\n### Step 0: Expert knowledge — non-obvious Service Quotas behaviors\n\nThese behaviors are easy to misjudge without operational Service Quotas\nexperience. Each changes a verdict if ignored:\n\n> Moved to [references\u002Fadvanced-patterns.md](references\u002Fadvanced-patterns.md#step-0-expert-knowledge--non-obvious-service-quotas-behaviors).\n> Fifteen diagnosis-changing behaviors: applied-vs-default, UsageMetric presence, exact CW dimensions, GlobalQuota scope, pending-vs-applied, request lifecycle, Adjustable:false, ingestion delay, no utilization field, API throttling, auto-case routing, pending-request quota, QuotaCode stability, Trusted Advisor deprecation, instant spikes.\n\n### Step 1: Utilization threshold — APPROACHING_LIMIT\n\nCompute utilization: `utilization_pct = (currentUsage \u002F appliedValue) * 100`.\n\nIf `utilization_pct >= 80`, the verdict is **APPROACHING_LIMIT**. This is\nthe highest-priority finding regardless of alarm state, UsageMetric\npresence, or increase request status. The operational risk of imminent\nexhaustion dominates all monitoring and structural concerns.\n\n**Boundary rule (critical):** the threshold is **inclusive** — `>= 80%`\nmeans a quota at exactly 80.0% utilization IS APPROACHING_LIMIT, not OK.\nFor example, 40\u002F50 VPCs (exactly 80%) is APPROACHING_LIMIT. Models and\noperators frequently misread `>=` as `>` (strictly greater than);\nthis is the single most common classification error. When in doubt, a\nquota at the exact boundary has no headroom for even one additional\nresource — treat it as approaching.\n\n- At exactly 80%: no headroom remains for growth. APPROACHING_LIMIT.\n- At 80-89%: immediate action required. Request an increase if\n  adjustable, or reduce usage.\n- At 90-100%: critical — any new resource creation or API burst may\n  fail with `LimitExceededException`.\n- At > 100%: the quota has been exceeded. AWS may allow brief overages\n  for some quotas but will block new operations imminently.\n\nIf a PENDING increase request exists, note it in FINDINGS but do NOT\ndowngrade the verdict — the applied quota has not changed yet.\n\nIf an APPROVED increase exists that has not yet propagated (applied\nvalue still shows old number), note the expected new value in FINDINGS.\n\n### Step 2: Structural config gap — CONFIG_GAP\n\nIf Step 1 did not match (utilization \u003C 80%), check for structural issues\nin this order:\n\n**(a) No UsageMetric defined.** The quota cannot be auto-monitored via\nCloudWatch. This is a CONFIG_GAP regardless of utilization — you have no\nprogrammatic signal to detect exhaustion. The remediation is to track\nmanually (enumerate live resources and compare against the applied value\non a schedule) or use AWS Config rules.\n\n**(b) `Adjustable: false` AND utilization >= 50%.** The quota is a hard\nlimit that cannot be increased, and utilization is high enough to be a\nfuture risk. This is a CONFIG_GAP because there is no increase path —\nthe remediation is architectural.\n\n**(c) Increase request DENIED or NOT_APPROVED AND utilization >= 50%.**\nA previous increase attempt was rejected and utilization is moderate-to-high.\nThe team has not addressed the underlying capacity issue. This is a\nCONFIG_GAP — either restructure usage or open a support case with\njustification.\n\n**(d) At AWS default, `Adjustable: true`, utilization >= 50%, AND the\nincrease request history is completely empty (zero entries of ANY\nstatus — no PENDING, no APPROVED, no DENIED, no NOT_APPROVED).**\nThe quota has never been increased and no request has ever been\nsubmitted, but utilization is high enough to warrant proactive action.\n\n**Disambiguation (critical for D8):** Step 2d matches ONLY when the\nrequest history list is truly empty. If there is ANY entry in the\nhistory — even a PENDING request or a DENIED request — Step 2d does\nNOT apply. A PENDING request means the team is actively seeking an\nincrease (skip to Step 3). A DENIED request is handled by Step 2c.\nOnly a completely empty history combined with utilization >= 50% at\nthe AWS default value triggers this CONFIG_GAP.\n\nIf none of (a)-(d) match, proceed to Step 3.\n\n### Step 3: Alarm coverage — NO_ALARM\n\nIf the quota has a `UsageMetric` (auto-trackable), utilization is \u003C 80%,\nand Step 2 found no structural issues, check for a CloudWatch alarm.\n\nIf NO CloudWatch alarm is configured on the `AWS\u002FUsage` metric for this\nquota, the verdict is **NO_ALARM**. The quota is trackable but no\nalerting exists — a sudden utilization spike would go undetected until\noperations fail.\n\nVerification: `aws cloudwatch describe-alarms --metric-name \u003CMetricName>\n--namespace AWS\u002FUsage --dimensions \u003Cdims from UsageMetric>`. An alarm\nis considered to exist ONLY when ALL of these conditions are met:\n(1) `MetricName` matches the `UsageMetric.MetricName` value, (2) every\ndimension from the `UsageMetric.Dimensions` list appears in the alarm\nwith the same name AND value, and (3) the alarm `Threshold` is at or\nbelow the applied quota value — an alarm threshold of 999999 for a\nquota of 50 is effectively disabled and should be treated as NO_ALARM.\nAn empty `describe-alarms` response or any dimension mismatch means\nNO_ALARM.\n\nNote: this step is only reached if Step 2 did not match. A quota at\ndefault with 60% utilization is CONFIG_GAP (Step 2d), not NO_ALARM — the\nstructural issue (should have requested an increase) takes precedence\nover the monitoring gap (no alarm).\n\n### Step 4: OK\n\nIf none of Steps 1-3 matched, the verdict is **OK**:\n- Utilization is below 80% (no approaching-limit risk).\n- No structural issues (Step 2 clean).\n- Either a CloudWatch alarm is configured, or the quota is low-risk\n  (low utilization, not adjustable, no metric).\n\nNo remediation required. Recommend periodic re-audit as usage grows.\n\n### Step 5: Aggregation — worst finding wins\n\nThe final verdict is the **maximum severity** across all findings, where\nAPPROACHING_LIMIT > CONFIG_GAP > NO_ALARM > OK:\n\n```text\nverdict = max(utilization_severity, config_gap_severity, alarm_severity)\n```\n\nIf no findings (all dimensions clean), the verdict is **OK**.\n\n**Creative audit extensions:** The four verdicts cover the standard\naudit dimensions. You MAY add supplementary FINDINGS lines for\nquota-specific risks the standard steps do not capture (e.g., a\nquota shared across instance types where one type is near exhaustion,\nor a quota whose growth rate has accelerated in the past 30 days).\nKeep the VERDICT to the four standard values — additional context\ngoes in FINDINGS, not in new verdict labels.\n\n## Output format (per quota)\n\n```text\nQUOTA: \u003Cquota-code> (\u003Cquota-name>)\nSERVICE: \u003Cservice-code>\nVERDICT: APPROACHING_LIMIT | NO_ALARM | CONFIG_GAP | OK\nREASON: \u003C1-2 sentences citing the worst finding and its step number>\nUTILIZATION: \u003Ccurrent>\u002F\u003Capplied> (\u003Cpct>%) [default: \u003Cdefault>] [unit: \u003Cunit>]\nFINDINGS:\n  - [APPROACHING_LIMIT] \u003Cfinding description (Step N)>\n  - [CONFIG_GAP] \u003Cfinding description (Step N)>\nREMEDIATION: \u003Cspecific CLI per finding, or \"None required\" if OK>\n```\n\n### Worked example — approaching limit with denied increase\n\n```text\nQUOTA: L-1212C26A (Running On-Demand Standard instances)\nSERVICE: ec2\nVERDICT: APPROACHING_LIMIT\nREASON: Utilization is at 85% of the applied quota (850\u002F1000 vCPUs).\nA prior increase request to 2000 vCPUs was DENIED — the applied value\nremains 1000 (Step 1).\nUTILIZATION: 850\u002F1000 (85%) [default: 1000] [unit: vCPUs]\nFINDINGS:\n  - [APPROACHING_LIMIT] 850\u002F1000 vCPUs (85%) — any new instance launch may fail (Step 1)\n  - [CONFIG_GAP] Increase request to 2000 DENIED on 2025-06-15 — no alternative capacity plan in place (Step 2c)\nREMEDIATION:\n  1. Open an AWS support case with workload justification for a vCPU increase.\n  2. Switch non-critical workloads to Spot Instances (separate quota).\n  3. Distribute across regions — the us-east-1 quota does not apply to us-west-2.\n```\n\n## Anti-Patterns — NEVER\n\n- NEVER compute utilization against the AWS default value instead of the\n  applied value. `list-service-quotas` returns the APPLIED value (which\n  may include approved increases). Using the default inflates utilization\n  and produces false-positive APPROACHING_LIMIT verdicts.\n\n- NEVER assume all quotas have a CloudWatch metric. Quotas without a\n  `UsageMetric` block CANNOT be auto-monitored. There is no\n  `AWS\u002FUsage` metric to alarm on — the only option is manual resource\n  enumeration. Treating \"no alarm\" as NO_ALARM for a non-trackable quota\n  is a category error; it is CONFIG_GAP (Step 2a).\n\n- NEVER treat a PENDING increase request as if it has taken effect. The\n  applied quota remains the old value until the request transitions to\n  APPROVED. Computing utilization against the requested future value\n  masks the current risk. Always use `get-service-quota` for the\n  authoritative applied value.\n\n- NEVER suggest `request-service-quota-increase` on an\n  `Adjustable: false` quota. It returns `ValidationException`. These are\n  hard limits — the only remediation is architectural (redistribute,\n  reduce, or switch services).\n\n- NEVER create a CloudWatch alarm on `AWS\u002FUsage` without copying the\n  exact dimensions from the `UsageMetric` block. Wrong dimension values\n  produce an alarm that monitors nothing — `DescribeAlarms` shows OK\n  while the actual metric has no data. Copy `Service`, `Resource`,\n  `Type`, and any `Class` dimension verbatim.\n\n- NEVER assume a regional quota increase applies across regions. A VPC\n  quota increase in us-east-1 does NOT apply to eu-west-1. Each region\n  has independent quotas for non-global quotas. Request increases\n  per-region.\n\n- NEVER use `QuotaName` as a stable identifier in automation.\n  `QuotaName` can be renamed by the service team; `QuotaCode` (e.g.,\n  `L-1212C26A`) is the stable identifier. Scripts that key on\n  `QuotaName` break silently when AWS renames a quota.\n\n- NEVER compare utilization across different `Unit` values. A quota of\n  \"100 Gigabytes\" and usage of \"50 Count\" is a data error, not 50%\n  utilization. Verify the unit before computing the ratio.\n\n- NEVER classify a quota at exactly 80.0% utilization as OK, NO_ALARM, or\n  CONFIG_GAP. The `>= 80%` threshold is **inclusive** — 40\u002F50 (exactly\n  80%) IS APPROACHING_LIMIT. The most common classification error is\n  treating the boundary as exclusive (`> 80%`). If utilization computes\n  to exactly 80.0%, the verdict MUST be APPROACHING_LIMIT, not OK.\n\n- NEVER ignore the 50% threshold for CONFIG_GAP. A quota at 55%\n  utilization that is adjustable, at default, and has no increase request\n  is a CONFIG_GAP — the team should have acted proactively. Waiting until\n  80% to flag it (where it becomes APPROACHING_LIMIT) leaves no\n  remediation window.\n\n- NEVER recommend deleting resources as the only remediation for an\n  APPROACHING_LIMIT quota. First check if the quota is adjustable and\n  submit an increase request. Resource cleanup is a complement, not a\n  substitute — the underlying capacity ceiling remains.\n\n- NEVER rely on Trusted Advisor service-limit checks as the sole\n  utilization signal. Trusted Advisor covers only a subset of quotas and\n  is deprecated for Business\u002FEnterprise support tiers. The modern\n  approach is Service Quotas + CloudWatch alarms on `AWS\u002FUsage`.\n\n- NEVER drain only the first page of `list-service-quotas` (100 per\n  page). The quota you need may be on page 2. Always iterate\n  `--starting-token` to completion.\n\n- NEVER exceed the concurrent pending-request quota. Each account has\n  a fixed limit on simultaneous PENDING increase requests per region\n  (typically 4 per region per service). When this limit is full,\n  `request-service-quota-increase` fails with `QuotaExceededException`\n  even though the target quota has ample headroom. Before submitting a\n  new request, check for stale PENDING entries via\n  `list-requested-service-quota-change-history --status PENDING` and\n  contact AWS support to withdraw expired ones — there is no public API\n  to programmatically close a pending request.\n\n- NEVER classify a quota with `UsageMetric` present, an alarm configured,\n  low utilization, and no structural issues as anything other than OK.\n  Over-classifying clean quotas erodes trust in the audit.\n\n## Pre-flight safety checks (run before any remediation CLI)\n\n- **MANDATORY CONFIRMATION GATE.** Before any state-changing operation\n  (`request-service-quota-increase`, closing a request), the auditor\n  MUST emit:\n  `CONFIRM: About to request a quota increase for \u003Cquota-code> on\n  \u003Cservice-code> from \u003Ccurrent> to \u003Cdesired>. This opens a request that\n  may require AWS support review. Proceed? (yes\u002Fno)`\n  Do NOT execute the CLI command until the operator confirms.\n- Before requesting an increase, verify the quota is `Adjustable: true`.\n  A request on a non-adjustable quota returns `ValidationException`.\n- Before requesting an increase, verify `--desired-value` is strictly\n  greater than the current applied value (`get-service-quota`). Equal or\n  lower values are rejected.\n- Check for existing PENDING requests on the same quota to avoid\n  duplicates: `aws service-quotas\n  list-requested-service-quota-change-history-by-quota --service-code\n  \u003Ccode> --quota-code \u003Ccode --status PENDING`.\n- Before creating a CloudWatch alarm, verify the `UsageMetric` dimensions\n  by querying the metric first:\n  `aws cloudwatch get-metric-statistics --namespace AWS\u002FUsage --metric-name\n  \u003CMetricName> --dimensions \u003Cdims> --start-time \u003Ciso> --end-time \u003Ciso>\n  --period 300 --statistics \u003CStatisticType>`. No data points = wrong\n  dimensions.\n- For regional quotas, confirm the correct `--region` is set on every\n  CLI call. A quota increase in the wrong region is a waste of a request\n  slot and cannot be transferred.\n\n## Remediation guidance\n\n### For APPROACHING_LIMIT\n\n1. If `Adjustable: true`, request an increase immediately:\n   `aws service-quotas request-service-quota-increase --service-code\n   \u003Ccode> --quota-code \u003Ccode> --desired-value \u003Cvalue>`\n   Use a desired value with at least 50% headroom above current usage\n   (e.g., if at 850\u002F1000, request 1500-2000).\n2. If a PENDING request already exists, note the expected resolution\n   time. Do NOT submit a duplicate — it wastes a request slot.\n3. If `Adjustable: false`, reduce usage immediately: terminate\n   non-critical resources, switch to Spot\u002Falternative instance types,\n   redistribute across accounts or regions.\n4. Create or verify a CloudWatch alarm at 80% and 90% thresholds.\n\n### For CONFIG_GAP\n\n- **No UsageMetric (Step 2a):** set up a scheduled audit (EventBridge +\n  Lambda) that enumerates live resources and compares against the\n  applied quota. There is no CloudWatch metric to alarm on.\n- **Adjustable: false (Step 2b):** document the architectural constraint\n  and plan capacity around it. Consider multi-account or multi-region\n  distribution.\n- **Denied increase (Step 2c):** open an AWS support case with\n  workload justification. Reduce usage in the interim.\n- **At default, no request (Step 2d):** submit a proactive increase\n  request. Even if current usage is moderate, growth will eventually\n  exhaust the default.\n\n### For NO_ALARM\n\nCreate a CloudWatch alarm on the `AWS\u002FUsage` metric:\n```bash\naws cloudwatch put-metric-alarm \\\n  --alarm-name \"quota-\u003Cquota-code>-80pct\" \\\n  --namespace AWS\u002FUsage \\\n  --metric-name \u003CMetricName from UsageMetric> \\\n  --dimensions \u003Cdims from UsageMetric> \\\n  --statistic \u003CStatisticType from UsageMetric> \\\n  --period 300 \\\n  --threshold \u003Capplied * 0.8> \\\n  --comparison-operator GreaterThanThreshold \\\n  --evaluation-periods 1 \\\n  --alarm-actions \u003Csns-topic-arn>\n```\n\n### For OK\n\nNo remediation required. Recommend:\n- Periodic re-audit as usage grows (monthly or per-deploy).\n- Verify the CloudWatch alarm threshold is still appropriate if the\n  applied quota changes.\n- For regional quotas, audit each region where the workload operates.\n\n## Reference: AWS\u002FUsage metric structure\n\nThe `UsageMetric` block maps to a CloudWatch metric in `AWS\u002FUsage`:\n\n| Field | Typical value | Notes |\n|---|---|---|\n| `MetricNamespace` | `AWS\u002FUsage` | Always this value. |\n| `MetricName` | `ResourceCount` | Sometimes `ResourceLimit` or service-specific. |\n| `Dimensions` | `Service`, `Resource`, `Type`, optionally `Class` | Copy verbatim — wrong values produce a silent alarm. |\n| `StatisticType` | `Sum` or `Maximum` | `Sum` for cumulative resources; `Maximum` for peak-based quotas. |\n\n**Organizations note:** the management account can view member-account\nquotas via `list-service-quotas --account-id \u003Cid>` but CANNOT submit\nincrease requests on their behalf — each member account must submit\nits own request.\n\n## Recent AWS features (2024-2026)\n\n> Moved to [references\u002Fadvanced-patterns.md](references\u002Fadvanced-patterns.md#recent-aws-features-2024-2026).\n> Organization-level quota management, more trackable quotas, increase-request automation.\n\n## References (load on demand)\n\n- [advanced-patterns](references\u002Fadvanced-patterns.md) — Step 0 expert knowledge (15 non-obvious Service Quotas behaviors) and recent AWS features\n\n## Domain\n\nAWS CloudOps \u002F Service Quotas Capacity Management & Monitoring.\n\n## AWS documentation\n\n- **Service documentation** — https:\u002F\u002Fdocs.aws.amazon.com\u002Fservicequotas\u002Flatest\u002Fuserguide\u002Fintro.html\n- **Security** — https:\u002F\u002Fdocs.aws.amazon.com\u002Fservicequotas\u002Flatest\u002Fuserguide\u002Fsecurity.html\n- **API reference** — https:\u002F\u002Fdocs.aws.amazon.com\u002Fservicequotas\u002F2019-06-24\u002Fapireference\u002F\n- **CLI reference** — https:\u002F\u002Fdocs.aws.amazon.com\u002Fcli\u002Flatest\u002Freference\u002Fservice-quotas\u002F\n","# Service Quotas Usage Auditor\n\n## Mindset\n\n**One-line takeaway:** the verdict is the **worst** finding across all\ndimensions, and three Service Quotas behaviors are easy to misjudge —\n**applied vs default** (list-service-quotas returns the APPLIED value,\nnot the AWS default), **UsageMetric presence** (quotas without a metric\nblock CANNOT be auto-monitored via CloudWatch), and **request status\nlifecycle** (a PENDING increase has not taken effect — the applied quota\nis still the old value until APPROVED).\n\nA quota is the hard ceiling on a specific AWS resource or API rate.\nExhausting one is a **silent outage**: API calls return\n`ThrottlingException` or `LimitExceededException` with no remediation\nwindow. Unlike a misconfigured policy (fixable by editing text), a quota\nthat blocks operations can stall an entire deployment pipeline while a\nsupport case grinds through the approval process.\n\n## Quick reference — verdict thresholds\n\n| Condition | Verdict | Step |\n|---|---|---|\n| Utilization >= 80% of applied quota (inclusive: exactly 80% counts) | **APPROACHING_LIMIT** | 1 |\n| `Adjustable: false` AND utilization >= 50% | **CONFIG_GAP** | 2b |\n| Increase request DENIED\u002FNOT_APPROVED AND utilization >= 50% | **CONFIG_GAP** | 2c |\n| At AWS default, `Adjustable: true`, utilization >= 50%, request history completely empty | **CONFIG_GAP** | 2d |\n| No `UsageMetric` block defined | **CONFIG_GAP** | 2a |\n| Has `UsageMetric`, utilization \u003C 80%, no CloudWatch alarm | **NO_ALARM** | 3 |\n| Utilization \u003C 50%, alarm in place or N\u002FA | **OK** | 4 |\n\nSee the ordered steps below for edge cases. Deep Service Quotas +\nCloudWatch integration details are in the\n[Deep reference](#deep-reference-service-quotas--cloudwatch-internals)\nsection at the end.\n\n## Pre-flight: quota metadata gate (run before classification)\n\nBefore evaluating utilization, classify the quota metadata itself.\nSeveral attributes **short-circuit** the audit — misclassifying them\nproduces false positives.\n\n| Attribute | Value | Effect on audit |\n|---|---|---|\n| `UsageMetric` | absent | **No CloudWatch metric exists.** Cannot auto-monitor utilization. This is a CONFIG_GAP (Step 2a) regardless of current utilization — you are flying blind. |\n| `UsageMetric` | present | Proceed — utilization is auto-trackable via `AWS\u002FUsage`. |\n| `Adjustable` | `false` | **Hard limit.** Cannot request an increase via `request-service-quota-increase`. If utilization >= 50%, the only remediation is architectural (distribute load, use a different service, reduce usage). Flag as CONFIG_GAP (Step 2b). |\n| `Adjustable` | `true` | Quota is increasable. Proceed to check increase request history. |\n| `GlobalQuota` | `true` | **Account-wide quota** (e.g., IAM users, S3 buckets). Applies across all regions. An increase applies globally. |\n| `GlobalQuota` | `false` | **Regional quota** (e.g., VPCs per region, EC2 vCPUs per region). An increase in us-east-1 does NOT apply to eu-west-1 — request per-region. |\n| `Unit` | `Gigabytes`, `Mbps`, etc. | **Unit mismatch risk.** Never compare a utilization value in one unit against a quota in another. A quota of \"100 Gigabytes\" at \"50 Count\" usage is a data error. |\n| Applied == Default | — | Quota has never been increased. Not inherently bad, but combined with utilization >= 50% and `Adjustable: true`, it signals that no proactive capacity management has occurred (Step 2d). |\n| Applied > Default | — | An increase was APPROVED at some point — active capacity management. This is a positive signal. |\n\n**Multi-service \u002F account-wide sweep note (pagination):** when auditing\nevery quota for a service, `aws service-quotas list-service-quotas\n--service-code \u003Ccode>` returns at most 100 per page. Use\n`--starting-token` from the prior `NextToken` to page through all\nquotas. The same applies to\n`list-requested-service-quota-change-history`. Always drain the\npaginator to completion — the long tail often contains the quota that\nmatters.\n\n**Malformed input handling.** If the quota snapshot is missing the\napplied value, missing quota code, has a utilization value not\nparseable as a number, or has a unit mismatch between utilization and\nquota, output an ERROR block and refuse classification. Do NOT guess\nthe utilization percentage — a wrong percentage produces a wrong\nverdict. Output:\n\n```text\nQUOTA: \u003Cquota-code or name>\nVERDICT: ERROR\nREASON: Quota snapshot is incomplete or malformed — cannot classify.\nREMEDIATION: Retrieve the canonical data with `aws service-quotas\nget-service-quota --service-code \u003Ccode> --quota-code \u003Ccode> --output json` and re-audit.\n```\n\n## Process — Classification logic (apply in order, aggregate worst)\n\n### Step 0: Expert knowledge — non-obvious Service Quotas behaviors\n\nThese behaviors are easy to misjudge without operational Service Quotas\nexperience. Each changes a verdict if ignored:\n\n> Moved to [references\u002Fadvanced-patterns.md](references\u002Fadvanced-patterns.md#step-0-expert-knowledge--non-obvious-service-quotas-behaviors).\n> Fifteen diagnosis-changing behaviors: applied-vs-default, UsageMetric presence, exact CW dimensions, GlobalQuota scope, pending-vs-applied, request lifecycle, Adjustable:false, ingestion delay, no utilization field, API throttling, auto-case routing, pending-request quota, QuotaCode stability, Trusted Advisor deprecation, instant spikes.\n\n### Step 1: Utilization threshold — APPROACHING_LIMIT\n\nCompute utilization: `utilization_pct = (currentUsage \u002F appliedValue) * 100`.\n\nIf `utilization_pct >= 80`, the verdict is **APPROACHING_LIMIT**. This is\nthe highest-priority finding regardless of alarm state, UsageMetric\npresence, or increase request status. The operational risk of imminent\nexhaustion dominates all monitoring and structural concerns.\n\n**Boundary rule (critical):** the threshold is **inclusive** — `>= 80%`\nmeans a quota at exactly 80.0% utilization IS APPROACHING_LIMIT, not OK.\nFor example, 40\u002F50 VPCs (exactly 80%) is APPROACHING_LIMIT. Models and\noperators frequently misread `>=` as `>` (strictly greater than);\nthis is the single most common classification error. When in doubt, a\nquota at the exact boundary has no headroom for even one additional\nresource — treat it as approaching.\n\n- At exactly 80%: no headroom remains for growth. APPROACHING_LIMIT.\n- At 80-89%: immediate action required. Request an increase if\n  adjustable, or reduce usage.\n- At 90-100%: critical — any new resource creation or API burst may\n  fail with `LimitExceededException`.\n- At > 100%: the quota has been exceeded. AWS may allow brief overages\n  for some quotas but will block new operations imminently.\n\nIf a PENDING increase request exists, note it in FINDINGS but do NOT\ndowngrade the verdict — the applied quota has not changed yet.\n\nIf an APPROVED increase exists that has not yet propagated (applied\nvalue still shows old number), note the expected new value in FINDINGS.\n\n### Step 2: Structural config gap — CONFIG_GAP\n\nIf Step 1 did not match (utilization \u003C 80%), check for structural issues\nin this order:\n\n**(a) No UsageMetric defined.** The quota cannot be auto-monitored via\nCloudWatch. This is a CONFIG_GAP regardless of utilization — you have no\nprogrammatic signal to detect exhaustion. The remediation is to track\nmanually (enumerate live resources and compare against the applied value\non a schedule) or use AWS Config rules.\n\n**(b) `Adjustable: false` AND utilization >= 50%.** The quota is a hard\nlimit that cannot be increased, and utilization is high enough to be a\nfuture risk. This is a CONFIG_GAP because there is no increase path —\nthe remediation is architectural.\n\n**(c) Increase request DENIED or NOT_APPROVED AND utilization >= 50%.**\nA previous increase attempt was rejected and utilization is moderate-to-high.\nThe team has not addressed the underlying capacity issue. This is a\nCONFIG_GAP — either restructure usage or open a support case with\njustification.\n\n**(d) At AWS default, `Adjustable: true`, utilization >= 50%, AND the\nincrease request history is completely empty (zero entries of ANY\nstatus — no PENDING, no APPROVED, no DENIED, no NOT_APPROVED).**\nThe quota has never been increased and no request has ever been\nsubmitted, but utilization is high enough to warrant proactive action.\n\n**Disambiguation (critical for D8):** Step 2d matches ONLY when the\nrequest history list is truly empty. If there is ANY entry in the\nhistory — even a PENDING request or a DENIED request — Step 2d does\nNOT apply. A PENDING request means the team is actively seeking an\nincrease (skip to Step 3). A DENIED request is handled by Step 2c.\nOnly a completely empty history combined with utilization >= 50% at\nthe AWS default value triggers this CONFIG_GAP.\n\nIf none of (a)-(d) match, proceed to Step 3.\n\n### Step 3: Alarm coverage — NO_ALARM\n\nIf the quota has a `UsageMetric` (auto-trackable), utilization is \u003C 80%,\nand Step 2 found no structural issues, check for a CloudWatch alarm.\n\nIf NO CloudWatch alarm is configured on the `AWS\u002FUsage` metric for this\nquota, the verdict is **NO_ALARM**. The quota is trackable but no\nalerting exists — a sudden utilization spike would go undetected until\noperations fail.\n\nVerification: `aws cloudwatch describe-alarms --metric-name \u003CMetricName>\n--namespace AWS\u002FUsage --dimensions \u003Cdims from UsageMetric>`. An alarm\nis considered to exist ONLY when ALL of these conditions are met:\n(1) `MetricName` matches the `UsageMetric.MetricName` value, (2) every\ndimension from the `UsageMetric.Dimensions` list appears in the alarm\nwith the same name AND value, and (3) the alarm `Threshold` is at or\nbelow the applied quota value — an alarm threshold of 999999 for a\nquota of 50 is effectively disabled and should be treated as NO_ALARM.\nAn empty `describe-alarms` response or any dimension mismatch means\nNO_ALARM.\n\nNote: this step is only reached if Step 2 did not match. A quota at\ndefault with 60% utilization is CONFIG_GAP (Step 2d), not NO_ALARM — the\nstructural issue (should have requested an increase) takes precedence\nover the monitoring gap (no alarm).\n\n### Step 4: OK\n\nIf none of Steps 1-3 matched, the verdict is **OK**:\n- Utilization is below 80% (no approaching-limit risk).\n- No structural issues (Step 2 clean).\n- Either a CloudWatch alarm is configured, or the quota is low-risk\n  (low utilization, not adjustable, no metric).\n\nNo remediation required. Recommend periodic re-audit as usage grows.\n\n### Step 5: Aggregation — worst finding wins\n\nThe final verdict is the **maximum severity** across all findings, where\nAPPROACHING_LIMIT > CONFIG_GAP > NO_ALARM > OK:\n\n```text\nverdict = max(utilization_severity, config_gap_severity, alarm_severity)\n```\n\nIf no findings (all dimensions clean), the verdict is **OK**.\n\n**Creative audit extensions:** The four verdicts cover the standard\naudit dimensions. You MAY add supplementary FINDINGS lines for\nquota-specific risks the standard steps do not capture (e.g., a\nquota shared across instance types where one type is near exhaustion,\nor a quota whose growth rate has accelerated in the past 30 days).\nKeep the VERDICT to the four standard values — additional context\ngoes in FINDINGS, not in new verdict labels.\n\n## Output format (per quota)\n\n```text\nQUOTA: \u003Cquota-code> (\u003Cquota-name>)\nSERVICE: \u003Cservice-code>\nVERDICT: APPROACHING_LIMIT | NO_ALARM | CONFIG_GAP | OK\nREASON: \u003C1-2 sentences citing the worst finding and its step number>\nUTILIZATION: \u003Ccurrent>\u002F\u003Capplied> (\u003Cpct>%) [default: \u003Cdefault>] [unit: \u003Cunit>]\nFINDINGS:\n  - [APPROACHING_LIMIT] \u003Cfinding description (Step N)>\n  - [CONFIG_GAP] \u003Cfinding description (Step N)>\nREMEDIATION: \u003Cspecific CLI per finding, or \"None required\" if OK>\n```\n\n### Worked example — approaching limit with denied increase\n\n```text\nQUOTA: L-1212C26A (Running On-Demand Standard instances)\nSERVICE: ec2\nVERDICT: APPROACHING_LIMIT\nREASON: Utilization is at 85% of the applied quota (850\u002F1000 vCPUs).\nA prior increase request to 2000 vCPUs was DENIED — the applied value\nremains 1000 (Step 1).\nUTILIZATION: 850\u002F1000 (85%) [default: 1000] [unit: vCPUs]\nFINDINGS:\n  - [APPROACHING_LIMIT] 850\u002F1000 vCPUs (85%) — any new instance launch may fail (Step 1)\n  - [CONFIG_GAP] Increase request to 2000 DENIED on 2025-06-15 — no alternative capacity plan in place (Step 2c)\nREMEDIATION:\n  1. Open an AWS support case with workload justification for a vCPU increase.\n  2. Switch non-critical workloads to Spot Instances (separate quota).\n  3. Distribute across regions — the us-east-1 quota does not apply to us-west-2.\n```\n\n## Anti-Patterns — NEVER\n\n- NEVER compute utilization against the AWS default value instead of the\n  applied value. `list-service-quotas` returns the APPLIED value (which\n  may include approved increases). Using the default inflates utilization\n  and produces false-positive APPROACHING_LIMIT verdicts.\n\n- NEVER assume all quotas have a CloudWatch metric. Quotas without a\n  `UsageMetric` block CANNOT be auto-monitored. There is no\n  `AWS\u002FUsage` metric to alarm on — the only option is manual resource\n  enumeration. Treating \"no alarm\" as NO_ALARM for a non-trackable quota\n  is a category error; it is CONFIG_GAP (Step 2a).\n\n- NEVER treat a PENDING increase request as if it has taken effect. The\n  applied quota remains the old value until the request transitions to\n  APPROVED. Computing utilization against the requested future value\n  masks the current risk. Always use `get-service-quota` for the\n  authoritative applied value.\n\n- NEVER suggest `request-service-quota-increase` on an\n  `Adjustable: false` quota. It returns `ValidationException`. These are\n  hard limits — the only remediation is architectural (redistribute,\n  reduce, or switch services).\n\n- NEVER create a CloudWatch alarm on `AWS\u002FUsage` without copying the\n  exact dimensions from the `UsageMetric` block. Wrong dimension values\n  produce an alarm that monitors nothing — `DescribeAlarms` shows OK\n  while the actual metric has no data. Copy `Service`, `Resource`,\n  `Type`, and any `Class` dimension verbatim.\n\n- NEVER assume a regional quota increase applies across regions. A VPC\n  quota increase in us-east-1 does NOT apply to eu-west-1. Each region\n  has independent quotas for non-global quotas. Request increases\n  per-region.\n\n- NEVER use `QuotaName` as a stable identifier in automation.\n  `QuotaName` can be renamed by the service team; `QuotaCode` (e.g.,\n  `L-1212C26A`) is the stable identifier. Scripts that key on\n  `QuotaName` break silently when AWS renames a quota.\n\n- NEVER compare utilization across different `Unit` values. A quota of\n  \"100 Gigabytes\" and usage of \"50 Count\" is a data error, not 50%\n  utilization. Verify the unit before computing the ratio.\n\n- NEVER classify a quota at exactly 80.0% utilization as OK, NO_ALARM, or\n  CONFIG_GAP. The `>= 80%` threshold is **inclusive** — 40\u002F50 (exactly\n  80%) IS APPROACHING_LIMIT. The most common classification error is\n  treating the boundary as exclusive (`> 80%`). If utilization computes\n  to exactly 80.0%, the verdict MUST be APPROACHING_LIMIT, not OK.\n\n- NEVER ignore the 50% threshold for CONFIG_GAP. A quota at 55%\n  utilization that is adjustable, at default, and has no increase request\n  is a CONFIG_GAP — the team should have acted proactively. Waiting until\n  80% to flag it (where it becomes APPROACHING_LIMIT) leaves no\n  remediation window.\n\n- NEVER recommend deleting resources as the only remediation for an\n  APPROACHING_LIMIT quota. First check if the quota is adjustable and\n  submit an increase request. Resource cleanup is a complement, not a\n  substitute — the underlying capacity ceiling remains.\n\n- NEVER rely on Trusted Advisor service-limit checks as the sole\n  utilization signal. Trusted Advisor covers only a subset of quotas and\n  is deprecated for Business\u002FEnterprise support tiers. The modern\n  approach is Service Quotas + CloudWatch alarms on `AWS\u002FUsage`.\n\n- NEVER drain only the first page of `list-service-quotas` (100 per\n  page). The quota you need may be on page 2. Always iterate\n  `--starting-token` to completion.\n\n- NEVER exceed the concurrent pending-request quota. Each account has\n  a fixed limit on simultaneous PENDING increase requests per region\n  (typically 4 per region per service). When this limit is full,\n  `request-service-quota-increase` fails with `QuotaExceededException`\n  even though the target quota has ample headroom. Before submitting a\n  new request, check for stale PENDING entries via\n  `list-requested-service-quota-change-history --status PENDING` and\n  contact AWS support to withdraw expired ones — there is no public API\n  to programmatically close a pending request.\n\n- NEVER classify a quota with `UsageMetric` present, an alarm configured,\n  low utilization, and no structural issues as anything other than OK.\n  Over-classifying clean quotas erodes trust in the audit.\n\n## Pre-flight safety checks (run before any remediation CLI)\n\n- **MANDATORY CONFIRMATION GATE.** Before any state-changing operation\n  (`request-service-quota-increase`, closing a request), the auditor\n  MUST emit:\n  `CONFIRM: About to request a quota increase for \u003Cquota-code> on\n  \u003Cservice-code> from \u003Ccurrent> to \u003Cdesired>. This opens a request that\n  may require AWS support review. Proceed? (yes\u002Fno)`\n  Do NOT execute the CLI command until the operator confirms.\n- Before requesting an increase, verify the quota is `Adjustable: true`.\n  A request on a non-adjustable quota returns `ValidationException`.\n- Before requesting an increase, verify `--desired-value` is strictly\n  greater than the current applied value (`get-service-quota`). Equal or\n  lower values are rejected.\n- Check for existing PENDING requests on the same quota to avoid\n  duplicates: `aws service-quotas\n  list-requested-service-quota-change-history-by-quota --service-code\n  \u003Ccode> --quota-code \u003Ccode --status PENDING`.\n- Before creating a CloudWatch alarm, verify the `UsageMetric` dimensions\n  by querying the metric first:\n  `aws cloudwatch get-metric-statistics --namespace AWS\u002FUsage --metric-name\n  \u003CMetricName> --dimensions \u003Cdims> --start-time \u003Ciso> --end-time \u003Ciso>\n  --period 300 --statistics \u003CStatisticType>`. No data points = wrong\n  dimensions.\n- For regional quotas, confirm the correct `--region` is set on every\n  CLI call. A quota increase in the wrong region is a waste of a request\n  slot and cannot be transferred.\n\n## Remediation guidance\n\n### For APPROACHING_LIMIT\n\n1. If `Adjustable: true`, request an increase immediately:\n   `aws service-quotas request-service-quota-increase --service-code\n   \u003Ccode> --quota-code \u003Ccode> --desired-value \u003Cvalue>`\n   Use a desired value with at least 50% headroom above current usage\n   (e.g., if at 850\u002F1000, request 1500-2000).\n2. If a PENDING request already exists, note the expected resolution\n   time. Do NOT submit a duplicate — it wastes a request slot.\n3. If `Adjustable: false`, reduce usage immediately: terminate\n   non-critical resources, switch to Spot\u002Falternative instance types,\n   redistribute across accounts or regions.\n4. Create or verify a CloudWatch alarm at 80% and 90% thresholds.\n\n### For CONFIG_GAP\n\n- **No UsageMetric (Step 2a):** set up a scheduled audit (EventBridge +\n  Lambda) that enumerates live resources and compares against the\n  applied quota. There is no CloudWatch metric to alarm on.\n- **Adjustable: false (Step 2b):** document the architectural constraint\n  and plan capacity around it. Consider multi-account or multi-region\n  distribution.\n- **Denied increase (Step 2c):** open an AWS support case with\n  workload justification. Reduce usage in the interim.\n- **At default, no request (Step 2d):** submit a proactive increase\n  request. Even if current usage is moderate, growth will eventually\n  exhaust the default.\n\n### For NO_ALARM\n\nCreate a CloudWatch alarm on the `AWS\u002FUsage` metric:\n```bash\naws cloudwatch put-metric-alarm \\\n  --alarm-name \"quota-\u003Cquota-code>-80pct\" \\\n  --namespace AWS\u002FUsage \\\n  --metric-name \u003CMetricName from UsageMetric> \\\n  --dimensions \u003Cdims from UsageMetric> \\\n  --statistic \u003CStatisticType from UsageMetric> \\\n  --period 300 \\\n  --threshold \u003Capplied * 0.8> \\\n  --comparison-operator GreaterThanThreshold \\\n  --evaluation-periods 1 \\\n  --alarm-actions \u003Csns-topic-arn>\n```\n\n### For OK\n\nNo remediation required. Recommend:\n- Periodic re-audit as usage grows (monthly or per-deploy).\n- Verify the CloudWatch alarm threshold is still appropriate if the\n  applied quota changes.\n- For regional quotas, audit each region where the workload operates.\n\n## Reference: AWS\u002FUsage metric structure\n\nThe `UsageMetric` block maps to a CloudWatch metric in `AWS\u002FUsage`:\n\n| Field | Typical value | Notes |\n|---|---|---|\n| `MetricNamespace` | `AWS\u002FUsage` | Always this value. |\n| `MetricName` | `ResourceCount` | Sometimes `ResourceLimit` or service-specific. |\n| `Dimensions` | `Service`, `Resource`, `Type`, optionally `Class` | Copy verbatim — wrong values produce a silent alarm. |\n| `StatisticType` | `Sum` or `Maximum` | `Sum` for cumulative resources; `Maximum` for peak-based quotas. |\n\n**Organizations note:** the management account can view member-account\nquotas via `list-service-quotas --account-id \u003Cid>` but CANNOT submit\nincrease requests on their behalf — each member account must submit\nits own request.\n\n## Recent AWS features (2024-2026)\n\n> Moved to [references\u002Fadvanced-patterns.md](references\u002Fadvanced-patterns.md#recent-aws-features-2024-2026).\n> Organization-level quota management, more trackable quotas, increase-request automation.\n\n## References (load on demand)\n\n- [advanced-patterns](references\u002Fadvanced-patterns.md) — Step 0 expert knowledge (15 non-obvious Service Quotas behaviors) and recent AWS features\n\n## Domain\n\nAWS CloudOps \u002F Service Quotas Capacity Management & Monitoring.\n\n## AWS documentation\n\n- **Service documentation** — https:\u002F\u002Fdocs.aws.amazon.com\u002Fservicequotas\u002Flatest\u002Fuserguide\u002Fintro.html\n- **Security** — https:\u002F\u002Fdocs.aws.amazon.com\u002Fservicequotas\u002Flatest\u002Fuserguide\u002Fsecurity.html\n- **API reference** — https:\u002F\u002Fdocs.aws.amazon.com\u002Fservicequotas\u002F2019-06-24\u002Fapireference\u002F\n- **CLI reference** — https:\u002F\u002Fdocs.aws.amazon.com\u002Fcli\u002Flatest\u002Freference\u002Fservice-quotas\u002F\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},"service-quotas, management, quota, limits, cloudwatch, monitoring, audit","2","Jacky Chan — AWS Community Builder","aws-cloudops","Management","0.1.0","Service Quotas, quota utilization, approaching limit, AWS\u002FUsage, CloudWatch alarm, quota increase, applied quota, default quota, Adjustable, UsageMetric, quota code, service limits, Trusted Advisor, quota monitoring, quota remediation","medium","false","Auditing AWS Service Quotas utilization, CloudWatch alarm coverage on AWS\u002FUsage metrics, and quota increase request history.","true","APPROACHING_LIMIT | NO_ALARM | CONFIG_GAP | OK","Input: either (a) a Service Quotas snapshot (quota code, service code, applied value, default value, UsageMetric, utilization, increase request history, CloudWatch alarm state), OR (b) a service-code for live-account audit via aws service-quotas list-service-quotas. Output: deterministic QUOTA\u002FVERDICT\u002FREASON\u002FUTILIZATION\u002FFINDINGS\u002F REMEDIATION block per quota, where VERDICT is one of APPROACHING_LIMIT, NO_ALARM, CONFIG_GAP, OK (or ERROR for malformed input).","audit service quotas, check quota utilization, approaching service limit, quota increase request, cloudwatch alarm on quota, service limits audit","Agent runtime that reads SKILL.md (Claude Code, Cursor, Windsurf, Codex, Gemini). No AWS CLI required for offline quota-snapshot classification. Live-account audits use aws service-quotas list-service-quotas, get-service-quota, get-aws-default-service-quota, list-requested-service-quota-change-history, and aws cloudwatch describe-alarbs-for-metric (AWS CLI v2, SSO or key-based credentials).","main","eeace51bd5dd1db2030729d83eeab988861553aa","skills\u002Fservice-quotas-usage-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 service-quotas-usage-auditor",{"tool":54,"command":55},"bunx","bunx skills add jhkchan\u002Faws-agent-skills --skill service-quotas-usage-auditor",{"tool":57,"command":58},"pnpm","pnpm dlx skills add jhkchan\u002Faws-agent-skills --skill service-quotas-usage-auditor","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Fjhkchan\u002Faws-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Fservice-quotas-usage-auditor\u002FSKILL.md",true,[62,68,72,75,79,83,87,91,95,99,103,107,111,115,119,123,127,131],{"path":63,"resourceType":64,"resourceTypeDesc":65,"sizeBytes":66,"sourceHtmlUrl":67},"eval\u002Ftest-cases.yaml","REFERENCE","参考文档",9891,"https:\u002F\u002Fgithub.com\u002Fjhkchan\u002Faws-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Fservice-quotas-usage-auditor\u002Feval\u002Ftest-cases.yaml",{"path":69,"resourceType":64,"resourceTypeDesc":65,"sizeBytes":70,"sourceHtmlUrl":71},"evals\u002Fbaselines\u002Fapproaching-limit-eighty-pct.md",667,"https:\u002F\u002Fgithub.com\u002Fjhkchan\u002Faws-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Fservice-quotas-usage-auditor\u002Fevals\u002Fbaselines\u002Fapproaching-limit-eighty-pct.md",{"path":73,"resourceType":64,"resourceTypeDesc":65,"sizeBytes":70,"sourceHtmlUrl":74},"evals\u002Fbaselines\u002Fdenied-increase-config-gap.md","https:\u002F\u002Fgithub.com\u002Fjhkchan\u002Faws-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Fservice-quotas-usage-auditor\u002Fevals\u002Fbaselines\u002Fdenied-increase-config-gap.md",{"path":76,"resourceType":64,"resourceTypeDesc":65,"sizeBytes":77,"sourceHtmlUrl":78},"evals\u002Fbaselines\u002Fhealthy-monitored-ok.md",548,"https:\u002F\u002Fgithub.com\u002Fjhkchan\u002Faws-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Fservice-quotas-usage-auditor\u002Fevals\u002Fbaselines\u002Fhealthy-monitored-ok.md",{"path":80,"resourceType":64,"resourceTypeDesc":65,"sizeBytes":81,"sourceHtmlUrl":82},"evals\u002Fbaselines\u002Fno-metric-cannot-track.md",627,"https:\u002F\u002Fgithub.com\u002Fjhkchan\u002Faws-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Fservice-quotas-usage-auditor\u002Fevals\u002Fbaselines\u002Fno-metric-cannot-track.md",{"path":84,"resourceType":64,"resourceTypeDesc":65,"sizeBytes":85,"sourceHtmlUrl":86},"evals\u002Fbaselines\u002Fpending-increase-still-high.md",654,"https:\u002F\u002Fgithub.com\u002Fjhkchan\u002Faws-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Fservice-quotas-usage-auditor\u002Fevals\u002Fbaselines\u002Fpending-increase-still-high.md",{"path":88,"resourceType":64,"resourceTypeDesc":65,"sizeBytes":89,"sourceHtmlUrl":90},"evals\u002Fbaselines\u002Ftrackable-missing-alarm.md",602,"https:\u002F\u002Fgithub.com\u002Fjhkchan\u002Faws-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Fservice-quotas-usage-auditor\u002Fevals\u002Fbaselines\u002Ftrackable-missing-alarm.md",{"path":92,"resourceType":64,"resourceTypeDesc":65,"sizeBytes":93,"sourceHtmlUrl":94},"evals\u002Feval_config.yaml",2080,"https:\u002F\u002Fgithub.com\u002Fjhkchan\u002Faws-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Fservice-quotas-usage-auditor\u002Fevals\u002Feval_config.yaml",{"path":96,"resourceType":64,"resourceTypeDesc":65,"sizeBytes":97,"sourceHtmlUrl":98},"evals\u002Fevals.json",5148,"https:\u002F\u002Fgithub.com\u002Fjhkchan\u002Faws-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Fservice-quotas-usage-auditor\u002Fevals\u002Fevals.json",{"path":100,"resourceType":64,"resourceTypeDesc":65,"sizeBytes":101,"sourceHtmlUrl":102},"evals\u002Fprompts\u002Fapproaching-limit-eighty-pct.md",856,"https:\u002F\u002Fgithub.com\u002Fjhkchan\u002Faws-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Fservice-quotas-usage-auditor\u002Fevals\u002Fprompts\u002Fapproaching-limit-eighty-pct.md",{"path":104,"resourceType":64,"resourceTypeDesc":65,"sizeBytes":105,"sourceHtmlUrl":106},"evals\u002Fprompts\u002Fdenied-increase-config-gap.md",926,"https:\u002F\u002Fgithub.com\u002Fjhkchan\u002Faws-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Fservice-quotas-usage-auditor\u002Fevals\u002Fprompts\u002Fdenied-increase-config-gap.md",{"path":108,"resourceType":64,"resourceTypeDesc":65,"sizeBytes":109,"sourceHtmlUrl":110},"evals\u002Fprompts\u002Fhealthy-monitored-ok.md",870,"https:\u002F\u002Fgithub.com\u002Fjhkchan\u002Faws-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Fservice-quotas-usage-auditor\u002Fevals\u002Fprompts\u002Fhealthy-monitored-ok.md",{"path":112,"resourceType":64,"resourceTypeDesc":65,"sizeBytes":113,"sourceHtmlUrl":114},"evals\u002Fprompts\u002Fno-metric-cannot-track.md",611,"https:\u002F\u002Fgithub.com\u002Fjhkchan\u002Faws-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Fservice-quotas-usage-auditor\u002Fevals\u002Fprompts\u002Fno-metric-cannot-track.md",{"path":116,"resourceType":64,"resourceTypeDesc":65,"sizeBytes":117,"sourceHtmlUrl":118},"evals\u002Fprompts\u002Fpending-increase-still-high.md",930,"https:\u002F\u002Fgithub.com\u002Fjhkchan\u002Faws-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Fservice-quotas-usage-auditor\u002Fevals\u002Fprompts\u002Fpending-increase-still-high.md",{"path":120,"resourceType":64,"resourceTypeDesc":65,"sizeBytes":121,"sourceHtmlUrl":122},"evals\u002Fprompts\u002Ftrackable-missing-alarm.md",774,"https:\u002F\u002Fgithub.com\u002Fjhkchan\u002Faws-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Fservice-quotas-usage-auditor\u002Fevals\u002Fprompts\u002Ftrackable-missing-alarm.md",{"path":124,"resourceType":64,"resourceTypeDesc":65,"sizeBytes":125,"sourceHtmlUrl":126},"examples\u002Fend-to-end.md",2937,"https:\u002F\u002Fgithub.com\u002Fjhkchan\u002Faws-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Fservice-quotas-usage-auditor\u002Fexamples\u002Fend-to-end.md",{"path":128,"resourceType":64,"resourceTypeDesc":65,"sizeBytes":129,"sourceHtmlUrl":130},"references\u002Fadvanced-patterns.md",7003,"https:\u002F\u002Fgithub.com\u002Fjhkchan\u002Faws-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Fservice-quotas-usage-auditor\u002Freferences\u002Fadvanced-patterns.md",{"path":132,"resourceType":64,"resourceTypeDesc":65,"sizeBytes":133,"sourceHtmlUrl":134},"skill.usf.yaml",4046,"https:\u002F\u002Fgithub.com\u002Fjhkchan\u002Faws-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Fservice-quotas-usage-auditor\u002Fskill.usf.yaml",[136,143,150,157,164,171],{"skillNo":137,"name":138,"nameZh":6,"description":139,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":140,"categories":141,"keywords":142,"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":144,"name":145,"nameZh":6,"description":146,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":147,"categories":148,"keywords":149,"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":151,"name":152,"nameZh":6,"description":153,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":154,"categories":155,"keywords":156,"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":158,"name":159,"nameZh":6,"description":160,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":161,"categories":162,"keywords":163,"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":165,"name":166,"nameZh":6,"description":167,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":168,"categories":169,"keywords":170,"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":172,"name":173,"nameZh":6,"description":174,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":175,"categories":176,"keywords":177,"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",[],[]]