[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M7ABPP7HJAVR0CMMMAC6C399":3,"skill-same:01M7ABPP7HJAVR0CMMMAC6C399":151},{"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":43,"commitSha":44,"skillDirectory":45,"forksCount":46,"repositoryHtmlUrl":47,"repositoryHomepageUrl":6,"ownerHtmlUrl":48,"ownerBlogUrl":6,"ownerDescription":49,"updatedAt":50,"install":51,"resources":64},"01M7ABPP7HJAVR0CMMMAC6C399","s3-batch-operations-operator",null,"Operates S3 Batch Operations at scale — job creation across operation types (copy, replace tag, restore from Glacier, replicate, invoke Lambda, put ACL, put object lock retention, put legal hold), manifest formats (S3 inventory report or CSV), completion reports, job priority and rate control (RequestsPerSecond), IAM permissions (batchoperations:* plus operation-specific grants), Lambda invoke for custom per-object processing, bulk storage class transition, bulk ACL fix, bulk replication backfill, bulk decrypt + re-encrypt with new KMS key, S3 Tables Batch Operations, and large-job handling for billion-object manifests. Runs deterministic pre-checks (manifest readability, report bucket writability, IAM chain, KMS decrypt\u002Fencrypt, Lambda invoke, rate-control) behind a CONFIRM gate and emits READY, BLOCKED, or COMPLETED per job. Use when running bulk object transformations, restoring Glacier archives at scale, backfilling replication, rotating KMS keys, or invoking a custom Lambda across billions of objects.","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\u002Fs3-batch-operations-operator\u002FSKILL.md",[],[],"2026-11-15 14:19:17",false,"---\nname: s3-batch-operations-operator\ndescription: Operates S3 Batch Operations at scale — job creation across operation types (copy, replace tag, restore from Glacier, replicate, invoke Lambda, put ACL, put object lock retention, put legal hold), manifest formats (S3 inventory report or CSV), completion reports, job priority and rate control (RequestsPerSecond), IAM permissions (batchoperations:* plus operation-specific grants), Lambda invoke for custom per-object processing, bulk storage class transition, bulk ACL fix, bulk replication backfill, bulk decrypt + re-encrypt with new KMS key, S3 Tables Batch Operations, and large-job handling for billion-object manifests. Runs deterministic pre-checks (manifest readability, report bucket writability, IAM chain, KMS decrypt\u002Fencrypt, Lambda invoke, rate-control) behind a CONFIRM gate and emits READY, BLOCKED, or COMPLETED per job. Use when running bulk object transformations, restoring Glacier archives at scale, backfilling replication, rotating KMS keys, or invoking a custom Lambda across billions of objects.\nlicense: Apache-2.0\ncompatibility: Agent runtime that reads SKILL.md (Claude Code, Cursor, Windsurf, Codex, Gemini). No AWS CLI required for offline plan classification. Live-account operations use aws s3control create-job, describe-job, update-job-priority, update-job-status, aws s3api list-bucket-inventory-configurations, get-bucket-location, get-bucket-versioning, aws lambda get-policy, aws iam list-attached-role-policies \u002F list-role-policies, aws kms describe-key \u002F get-key-policy, and aws logs filter-log-events (AWS CLI...\nmetadata:\n  domain: aws-cloudops\n  complexity: high\n  requires_llm: 'true'\n  phase: '4'\n  supports_pipeline: 'true'\n  entry_point: 'false'\n  family: Storage\n  task_type: operate\n  skill_class: capability\n  lifecycle_status: active\n  verdict_shape: READY | BLOCKED | COMPLETED\n  when_to_use: Creating an S3 Batch Operations job (copy, replace tag, restore from Glacier, replicate, invoke Lambda, put ACL, put object lock retention, put object lock legal hold), diagnosing a failed or stalled job, cancelling a runaway job, adjusting job priority or rate control, backfilling replication across existing objects, bulk-transitioning storage class, bulk-rotating KMS keys via decrypt + re-encrypt, invoking a custom Lambda across a manifest of objects, or planning a billion- object batch run with safe rate limits.\n  activation_triggers: S3 Batch Operations, batch copy objects, bulk restore Glacier, bulk replace tags, batch invoke Lambda, bulk replication backfill, bulk KMS re-encrypt, bulk storage class transition, put object lock retention batch, batch operations manifest, S3 inventory manifest, CSV manifest, completion report, RequestsPerSecond batch, cancel batch job, update job priority, S3 Tables batch operations, billion objects batch\n  invocation_schema: 'Input: either (a) a job specification (operation type, source bucket, manifest format and location, report bucket, IAM role ARN, operation- specific parameters, priority, rate control) for plan classification, OR (b) a JobId for live-account diagnose or cancel operations. Output: deterministic OPERATION \u002F VERDICT \u002F PRE_CHECKS \u002F STEPS \u002F POST_VERIFY \u002F NOTES block per job, where VERDICT is one of READY, BLOCKED, COMPLETED.'\n  version: 0.1.0\n  author: Jacky Chan — AWS Community Builder\n  keywords: S3 Batch Operations, create-job, manifest, S3 inventory report, CSV manifest, completion report, job priority, RequestsPerSecond, rate control, copy operation, replace tag, restore from Glacier, replicate operation, invoke Lambda, put ACL, object lock retention, object lock legal hold, bulk storage class transition, bulk ACL fix, bulk replication backfill, bulk KMS re-encrypt, S3 Tables, billion objects, batchoperations, s3:GetObject, s3:PutObject\n  tags: aws, s3, s3control, storage, batch, bulk, glacier, kms, lambda, object-lock, replication, operate\n---\n\n# S3 Batch Operations Operator\n\n## What this skill does\n\nExecutes S3 Batch Operations jobs correctly and safely at any scale. Runs\ndeterministic pre-checks before any state-changing CLI (manifest format\nand readability, report bucket writability, IAM role permission chain\nincluding operation-specific grants, KMS decrypt\u002Fencrypt path for\nSSE-KMS objects, Lambda invoke permission for Lambda operations,\nrate-control validity, manifest object-count estimate), executes the\n`create-job` behind a CONFIRM gate, and verifies the result by polling\n`describe-job` until `Status: Complete`, reading the completion report,\nand reconciling `Failed` count against an acceptable threshold. Every\njob plan names the operation-specific IAM grants; every diagnose-job\nsurfaces the failure-mode table so the operator knows whether the root\ncause is permissions, manifest format, throttling, or a per-object error.\n\n## Quick navigation\n\n| Section | What it covers | When to read it |\n|---|---|---|\n| **§ Quick reference** | Verdict thresholds (BLOCKED\u002FREADY\u002FCOMPLETED) + pre-check priority | Before any operation |\n| **§ Mindset** | Why a batch job is a manifest + role + report + rate-control quad, the false-green trap | Understanding the safety model |\n| **§ Pre-flight** | Job spec gate — manifest versioning, report bucket region, role ARN | Before executing any CLI |\n| **§ Process** | Per-operation planning: create-job, diagnose-job, cancel-job, update-priority | When choosing which operation to run |\n| **§ Output format** | STRICT output contract — OPERATION\u002FVERDICT\u002FPRE_CHECKS\u002FSTEPS\u002FPOST_VERIFY\u002FNOTES template | Formatting the response |\n| **§ Anti-Patterns** | NEVER list — common mistakes that silently fail jobs or skip objects | Review before risky operations |\n| **§ Pre-flight safety** | Capture pre-state, IAM role policy, KMS key policy, manifest validation | Defense-in-depth |\n| **§ Expert heuristic** | Batch Operations is async and per-object failures are silent — never trust `Active` alone | Avoiding the false-green trap |\n\n## STRICT output contract\n\nEVERY response MUST end with a single fenced text block in this exact\nshape (the operator's downstream tooling greps for it). No deviations:\n\n```text\nOPERATION: \u003Ccreate-job | diagnose-job | cancel-job | update-priority>\nVERDICT: READY | BLOCKED | COMPLETED\nTARGET: \u003Cmanifest-bucket\u002Fpath> (operation: \u003Ctype>, role: \u003Carn-or-\"none\">)\nPRE_CHECKS:\n  - [PASS] \u003Ccheck description>\n  - [FAIL] \u003Ccheck description> — \u003Creason>\nSTEPS:\n  1. \u003CCLI command with flags populated, or \"(none — pre-checks failed)\">\n  2. \u003Cwait \u002F monitoring command>\nPOST_VERIFY:\n  - [PASS] \u003Cverification description> | (pending execution)\n  - [FAIL] \u003Cverification description> — \u003Creason>\nNOTES: \u003Crate-control, monitoring, caveats>\n```\n\nRules of the contract:\n\n- VERDICT is exactly one of `READY`, `BLOCKED`, `COMPLETED`. No other\n  values. `READY` means pre-checks passed and a CONFIRM gate is\n  pending; `BLOCKED` means at least one pre-check failed — do NOT emit\n  STEPS that mutate state; `COMPLETED` means post-verification passed\n  after the job finished.\n- If PRE_CHECKS has any `[FAIL]`, VERDICT MUST be `BLOCKED` and STEPS\n  MUST be `(none — pre-checks failed)`.\n- If VERDICT is `READY`, STEPS[1] MUST be the CONFIRM prompt and\n  STEPS[2] MUST be the actual `create-job` CLI.\n- If VERDICT is `COMPLETED`, POST_VERIFY MUST contain at least one\n  `[PASS]` and no `[FAIL]`.\n- Never wrap the block in JSON, never abbreviate the field names,\n  never omit a section. If a section is empty, write `(none)`.\n\n## Quick reference — verdict thresholds\n\n| Verdict | Trigger condition | Action |\n|---|---|---|\n| `BLOCKED` | One or more pre-checks failed (manifest unreadable, report bucket not writable, role missing `batchoperations:*` or operation-specific grants, KMS decrypt denied for SSE-KMS source, Lambda resource policy missing `batchoperations.amazonaws.com` principal, rate-control invalid, manifest contains zero objects, destination bucket missing for copy) | List failures, do NOT execute |\n| `READY` | All pre-checks passed; awaiting CONFIRM gate | Emit exact `create-job` CLI with all flags populated, wait for operator yes |\n| `COMPLETED` | Job reached `Status: Complete` AND post-verification passed (`Failed` count within threshold, completion report readable, sample objects verified) | Emit verification results, monitoring plan |\n\n**Priority order for pre-checks (apply in this sequence, all must pass\nfor READY):**\n\n1. **Manifest format and readability.** Manifest is a versioned S3\n   inventory report (with `manifest.json` + `manifest.checksum`) OR a\n   CSV file (`bucket,key` per line, versioned CSV requires a third\n   column). The role can read the manifest object(s).\n2. **Report bucket configured and writable.** `ReportBucket` is in the\n   same Region as the Batch Operations job. Role has `s3:PutObject` on\n   the report prefix.\n3. **IAM role permission chain.** Role has `batchoperations:*` on the\n   job ARN (or `*`), plus read on the manifest bucket and write on the\n   target bucket(s). Operation-specific grants (see table below).\n4. **KMS decrypt path on source.** For SSE-KMS objects, the role has\n   `kms:Decrypt` on the source key ARN AND the source key policy\n   grants the role.\n5. **KMS encrypt path on destination.** For copy or re-encrypt\n   operations, the role has `kms:Encrypt` on the destination key ARN\n   AND the destination key policy grants the role.\n6. **Lambda invoke permission.** For invoke operations, the Lambda's\n   resource-based policy allows `batchoperations.amazonaws.com` to\n   `lambda:InvokeFunction`, scoped to the job role ARN.\n7. **Rate-control validity.** `RequestsPerSecond` (if set) is in\n   [1, 2500]. `CompletionWindow` (if set) is between 1 hour and 1 year.\n8. **Object-count estimate.** Manifest object count produces an ETA\n   within the configured `CompletionWindow`. For billion+ object\n   manifests, verify the rate limit produces a sub-window ETA.\n\n**Cost\u002Ftime baselines (2026):**\n\n- Per-object processing time: 0.5-2 seconds typical (operation-\n  dependent; Glacier restore and Lambda invoke are slower).\n- Default rate: S3 Batch Operations auto-scales within the account's\n  S3 request budget; explicit `RequestsPerSecond` caps it.\n- Job setup latency: 1-5 minutes from `create-job` to `Active`.\n- Completion report: written within minutes of job `Complete`.\n- Cost: per-request charges (PUT\u002FGET\u002Frestore) plus a Batch Operations\n  per-object processing fee (see S3 pricing).\n\n## Mindset\n\n**One-line takeaway:** `Status: Active` on a Batch Operations job is a\nscheduling claim, not a delivery claim. A job is only \"done\" when\n`Status: Complete`, the completion report shows the expected success\ncount, and a sample of objects verifies the operation actually applied.\nDriven by three S3 Batch Operations realities:\n\n- **A batch job is a manifest + role + report + rate-control quad.**\n  The manifest enumerates objects; the role grants the operation;\n  the report records results; the rate-control prevents runaway costs.\n  A single broken link makes the entire job fail or silently skip —\n  the job shows `Active` on dashboards while objects are untouched.\n- **Per-object failures are silent.** A job can complete with tens of\n  thousands of `Failed` entries while `Status: Complete` looks green.\n  Operators who do not read the completion report discover the failure\n  weeks later when downstream effects surface.\n- **Large jobs (billion+ objects) need explicit rate-control.** Without\n  `RequestsPerSecond`, the job can saturate the source or destination\n  bucket's request budget, crowd out application traffic, and exhaust\n  KMS request quotas (especially for decrypt+re-encrypt). Always set a\n  rate limit for jobs over 100 million objects.\n\n## Pre-flight: job spec gate\n\nRun before classification. Misclassifying these produces wrong plans.\n\n**Pagination:** `list-bucket-inventory-configurations` paginates at\n100. `describe-job` returns the full job descriptor in one call. CSV\nmanifests have no native pagination — estimate the object count from\nthe file size (each line averages 80-120 bytes for `bucket,key`).\n\n**Live-account pre-flight (skip if offline plan audit):**\nThe eight live-account pre-flight probes (manifest head-object, report-bucket region, versioning capture, role policy chain, KMS key state, destination KMS, Lambda policy, describe-job): [references\u002Fdiagnostic-commands.md](references\u002Fdiagnostic-commands.md).\n\n**Malformed input:** if the input JSON is invalid or missing required\nfields, emit `VERDICT: ERROR` with `REASON: Job spec is not valid JSON\nor is missing required fields — cannot plan.` and `REMEDIATION:\nRe-validate the spec against the operation schema in references\u002F\noperation-iam-matrix.md and re-plan.`\n\n| Job spec attribute | Effect on operation |\n|---|---|\n| `Manifest.Spec.Format: S3InventoryReport` | Requires `manifest.json` + `manifest.checksum`. Versioned inventory required for version-aware operations. |\n| `Manifest.Spec.Format: CSV` | Each line is `bucket,key` (unversioned) or `bucket,key,versionId` (versioned). No header row. |\n| `Operation` mismatch with manifest versioning | A versioned operation (e.g., object lock retention) on an unversioned manifest BLOCKS. |\n| `Report.Bucket` not in job Region | BLOCKED. Batch Operations requires the report bucket in the same Region as the job. |\n| `Role` missing `iam:PassRole` on the caller side | `create-job` itself fails with `AccessDenied`. Caller needs `iam:PassRole` on the role ARN. |\n| `RequestsPerSecond` outside [1, 2500] | BLOCKED. |\n| `Priority` outside [-2147483648, 2147483647] | BLOCKED. Higher value = higher priority. |\n| Manifest containing zero objects | BLOCKED. |\n| `ToggleEnabled: false` | Job created in `Suspended` state; will not run until `update-job-status` sets `Ready`. |\n\n## Process — operation planning (apply in order)\n\n### Step 0: Expert knowledge — non-obvious Batch Operations behaviors\n\nThese behaviors are easy to misjudge without operational experience.\nEach changes a plan if ignored:\nThe 12 non-obvious behaviors (two-sided role, versioned-inventory requirement, no CSV header row, fixed Lambda payload, two-stage Glacier restore, ServerSideCopy re-encrypt, full-tagset replace, global priority, CompletionWindow hint, report scope, S3 Tables operations, transient throttling): [references\u002Fadvanced-patterns.md](references\u002Fadvanced-patterns.md).\n\n\n### Step 1: Pre-check gate — BLOCKED if any check fails\n\nRun ALL pre-checks for the chosen operation. If ANY fails, the verdict\nis BLOCKED with the failed checks in PRE_CHECKS. Do NOT execute.\n\n**For ALL operations:**\n1. Manifest exists and is readable by the role.\n2. Manifest format matches the operation's versioning requirement.\n3. Report bucket exists in the same Region as the planned job.\n4. Caller has `s3control:CreateJob` AND `iam:PassRole` on the role.\n5. Role has `batchoperations:*` (or scoped equivalent).\n6. Role has read on the manifest bucket (`s3:GetObject` on the\n   manifest prefix).\n7. Role has `s3:PutObject` on the report bucket prefix.\n8. `RequestsPerSecond` (if set) is in [1, 2500].\n9. `Priority` is within the int32 range.\n10. Manifest object count is non-zero.\n\n**Operation-specific pre-checks:**\n\n| Operation | Additional checks |\n|---|---|\n| **Copy** | Destination bucket exists; role has `s3:PutObject` on destination; for SSE-KMS source, role has `kms:Decrypt`; for SSE-KMS destination, role has `kms:Encrypt` + destination key policy grants role. |\n| **ReplaceTag** | Role has `s3:PutObjectTagging` and `s3:DeleteObjectTagging` on the target. New tagset does not exceed 10 keys \u002F 256 chars per value. |\n| **Restore** | Source objects are in GLACIER \u002F GLACIER_IR \u002F DEEP_ARCHIVE. Role has `s3:RestoreObject`. `Days` and `GlacierJobTier` (Standard\u002FBulk) are set. |\n| **Replicate** | Source bucket has replication configured; destination rule exists. Role has `s3:ReplicateObject`, `s3:ReplicateDelete`. |\n| **Invoke** | Lambda `State: Active`. Lambda resource-based policy allows `batchoperations.amazonaws.com` to `lambda:InvokeFunction`. Lambda reserved concurrency >= `RequestsPerSecond`. Lambda timeout >= 60s (Batch retries on timeout, but a 3s timeout is almost always wrong). |\n| **PutACL** | Role has `s3:PutObjectAcl` on the target. Canned ACL or access-control policy is valid. |\n| **PutObjectLockRetention** | Target bucket has Object Lock enabled. Role has `s3:PutObjectRetention`. `Mode` is `GOVERNANCE` or `COMPLIANCE`. `RetainUntilDate` is in the future. |\n| **PutObjectLockLegalHold** | Target bucket has Object Lock enabled. Role has `s3:PutObjectLegalHold`. `Status` is `ON` or `OFF`. |\n\n**Job failure-mode table (use during diagnose-job):**\nThe nine-row failure-mode table mapping `describe-job` symptoms and completion-report failure codes to root causes and fixes: [references\u002Ferror-handling.md](references\u002Ferror-handling.md).\n\n### Step 2: READY — emit operation plan\n\nIf all pre-checks pass, emit `VERDICT: READY` with the exact CLI\nsequence and the CONFIRM gate. The plan includes:\n\n- The exact `aws s3control create-job` CLI with the JSON operation\n  spec inline or referenced from a file.\n- The expected duration (based on manifest size and rate limit).\n- The expected side-effects (object count processed, KMS request\n  volume, Lambda invocation count).\n- The CONFIRM gate prompt.\n- The monitoring commands (`describe-job` polling cadence, CloudWatch\n  metrics to watch).\n\n### Step 3: Execute behind CONFIRM gate\n\n- **MANDATORY CONFIRMATION GATE.** Before `create-job`,\n  `update-job-status`, `update-job-priority`, or any state-changing\n  CLI, emit: `CONFIRM: About to \u003Coperation> on job \u003Cid-or-\"new\"> in\n  account \u003Caccount> region \u003Cregion>. This will \u003Cconsequence>.\n  Proceed? (yes\u002Fno)`. Do NOT execute until the operator confirms.\n- Capture pre-state: if modifying an existing job,\n  `aws s3control describe-job --account-id \u003Caccount> --job-id \u003Cid>\n  --output json > \u002Ftmp\u002F\u003Cid>-pre-$(date +%s).json`.\n- For new jobs: write the operation spec to a local JSON file and\n  pass via `--operation`. Capture the returned `JobId`.\n- Poll `describe-job` every 60 seconds (or longer for billion-object\n  jobs) until `Status` transitions away from `Active`.\n\n### Step 4: Post-verification — COMPLETED\n\nAfter the job finishes, run post-verification. ALL checks must pass\nfor `COMPLETED`.\n\n1. `describe-job --job-id \u003Cid>` — confirm `Status: Complete`.\n2. Read `ProgressSummary` — `NumberOfTasksSucceeded` matches the\n   expected count; `NumberOfTasksFailed` is within the acceptable\n   threshold (typically 0; for billion-object jobs, \u003C 0.01%).\n3. Read the completion report (`Report.Bucket`). For `FailedTasksOnly`\n   scope, confirm the report is non-empty only if there were failures.\n4. Sample 5 objects from the manifest — verify the operation applied\n   (e.g., for copy, `head-object` on destination; for KMS re-encrypt,\n   verify `SSEKMSKeyId` matches the new key; for tag replace, verify\n   `get-object-tagging` matches the new tagset).\n5. For Glacier restore: sample 5 objects — verify `Restore` header\n   shows `ongoing-request=\"false\"` (or, for bulk tier, that the\n   request is queued).\n6. For Lambda invoke: check CloudWatch Logs for the Lambda — no\n   errors above the expected threshold.\n\nIf ANY verification fails, emit `VERDICT: ERROR` with the failure\ndetails — do not claim COMPLETED. A failed verification typically means\nthe job completed with unexpected failures; route to diagnose-job.\n\n## Output format (per operation)\n\nSee § STRICT output contract above. The block MUST appear exactly in\nthat shape at the end of the response.\n\n### Worked example — create-job (bulk KMS re-encrypt)\n\n```text\nOPERATION: create-job\nVERDICT: READY\nTARGET: s3:\u002F\u002Fprod-inventory\u002F2026-08-01\u002Fmanifest.json\n        (operation: copy with new SSE-KMS key, role:\n         arn:aws:iam::111111111111:role\u002FS3BatchOpsRole)\nPRE_CHECKS:\n  - [PASS] Manifest readable, format S3InventoryReport V2 (versioned)\n  - [PASS] Report bucket prod-batch-reports in us-east-1 (matches job)\n  - [PASS] Caller has s3control:CreateJob + iam:PassRole on\n    arn:aws:iam::111111111111:role\u002FS3BatchOpsRole\n  - [PASS] Role has batchoperations:*, s3:GetObject on prod-inventory,\n    s3:PutObject on prod-batch-reports\n  - [PASS] Role has s3:GetObject + s3:PutObject on prod-data\n  - [PASS] Role has kms:Decrypt on arn:aws:kms:us-east-1:111111111111:key\u002Fold-key\n  - [PASS] Role has kms:Encrypt on arn:aws:kms:us-east-1:111111111111:key\u002Fnew-key\n  - [PASS] New KMS key policy grants role Encrypt\n  - [PASS] RequestsPerSecond: 500 (in [1, 2500])\n  - [PASS] Manifest object count: 12,400,000 — ETA ~6.9h at 500 RPS\n    (within CompletionWindow: 24h)\nSTEPS:\n  1. CONFIRM: About to create a Batch Operations copy+re-encrypt job\n     on 12,400,000 objects in bucket prod-data (account\n     111111111111, region us-east-1), overwriting in place with new\n     SSE-KMS key arn:aws:kms:us-east-1:111111111111:key\u002Fnew-key, at\n     500 RPS. Estimated cost: ~$248 (PUT+GET+SSE-KMS). Proceed?\n     (yes\u002Fno)\n  2. aws s3control create-job \\\n       --account-id 111111111111 \\\n       --region us-east-1 \\\n       --priority 50 \\\n       --role-arn arn:aws:iam::111111111111:role\u002FS3BatchOpsRole \\\n       --operation '{\"S3InitiateRestoreObject\": null}' \\\n       --manifest '{\"Spec\":{\"Format\":\"S3InventoryReport\",\"Fields\":[\"Bucket\",\"Key\",\"VersionId\"]},\"Location\":{\"ObjectArn\":\"arn:aws:s3:::prod-inventory\u002F2026-08-01\u002Fmanifest.json\",\"ETag\":\"\u003Cetag>\"}}' \\\n       --report '{\"Bucket\":\"arn:aws:s3:::prod-batch-reports\",\"Prefix\":\"kms-reencrypt-2026-08\u002F\",\"Format\":\"Report_CSV_20180820\",\"ReportScope\":\"FailedTasksOnly\"}' \\\n       --client-request-token \"$(uuidgen)\"\n     (replace --operation with the actual copy-with-new-encryption JSON)\n  3. Poll: aws s3control describe-job --account-id 111111111111\n     --job-id \u003Creturned-id> --query 'Status'\n     (every 5 minutes for a 7-hour job)\nPOST_VERIFY:\n  - (pending execution)\nNOTES:\n  - Bulk KMS re-encrypt is a server-side copy with new SSE-KMS\n    metadata, overwriting in place. Set NewObjectMetadata.SSEKMSKeyId\n    to the new key in the operation JSON. Source ~12.4M Decrypt +\n    destination ~12.4M Encrypt requests. Verify KMS quota first.\n  - Use FailedTasksOnly report scope to avoid a multi-GB report.\n  - CloudWatch alarm on KMS Throttles during the job window.\n```\n\n### Worked example — diagnose-job (BLOCKED with remediation)\n\nFull diagnose-job BLOCKED example (Lambda reserved-concurrency root cause, remediation options, re-run of failed objects): [references\u002Fworked-examples.md](references\u002Fworked-examples.md).\n\n### Worked example — create-job (Glacier bulk restore, COMPLETED)\n\nFull Glacier bulk-restore COMPLETED example (two-stage restore verification): [references\u002Fworked-examples.md](references\u002Fworked-examples.md).\n\n## Anti-Patterns — NEVER\n\n- NEVER create a Batch Operations job without verifying the IAM role\n  has the operation-specific grants. `batchoperations:*` is the job-\n  control plane permission; the actual object operations need\n  `s3:GetObject`, `s3:PutObject`, `s3:RestoreObject`,\n  `s3:PutObjectTagging`, `s3:PutObjectRetention`, or\n  `s3:PutObjectAcl` depending on the operation type. Missing grants\n  produce a job that completes with 100% failure.\n\n- NEVER include a header row in a CSV manifest. The first line is\n  parsed as an object. A header `bucket,key` produces a `NoSuchKey`\n  failure for object `key` in bucket `bucket`. Use S3 inventory\n  format if you want structured metadata.\n\n- NEVER assume `Status: Complete` means all objects were successfully\n  processed. Read the completion report and reconcile\n  `NumberOfTasksFailed` against an acceptable threshold. For large\n  jobs, use `FailedTasksOnly` report scope and treat any non-empty\n  report as a follow-up item.\n\n- NEVER run a billion-object job without an explicit\n  `RequestsPerSecond`. Default auto-scaling can saturate the source\n  bucket, KMS quota, or Lambda concurrency, producing throttling\n  cascades that exhaust the retry budget.\n\n- NEVER assume Glacier restore is finished when the Batch Operations\n  job is `Complete`. The job submits restore requests; the actual\n  restore is asynchronous on the Glacier side. Poll `head-object\n  --restore` for `ongoing-request=\"false\"` before downstream use.\n\n- NEVER pass a role ARN at `create-job` without verifying the caller\n  has `iam:PassRole` on that ARN. The `create-job` API itself fails\n  with `AccessDenied` before any object processing.\n\n- NEVER set all jobs to `Priority: 2147483647`. Priority is a relative\n  scheduling signal within the account + Region. Inflating every job\n  to max nullifies the priority system and produces contention.\n\n- NEVER use the `Task` report scope for billion-object jobs. The\n  completion report will be a multi-GB CSV. Use `FailedTasksOnly` and\n  read the `ProgressSummary` for the success count.\n\n- NEVER attempt a `PutObjectLockRetention` operation on a bucket\n  without Object Lock enabled. The operation fails per-object with\n  `InvalidRequest`. Verify `get-object-lock-configuration` before\n  creating the job.\n\n- NEVER assume Lambda invocations are idempotent. Batch Operations\n  retries with the same `taskId`. If the Lambda has side effects\n  (writes to a database, sends a notification), it MUST deduplicate\n  on `taskId`.\n\n- NEVER trust `describe-job.ProgressSummary` alone for cost\n  verification. Cross-check S3 and KMS CloudWatch request metrics\n  during the job window — throttles and retries inflate actual\n  request count above the object count.\n\n- NEVER forget the manifest's `ETag` field. For S3 inventory\n  manifests, the `manifest.json` object has an ETag that must be\n  passed in the `Manifest.Location.ETag` field. Omitting it produces\n  a `400 MalformedXML` error from `create-job`.\n\n- NEVER auto-execute a state-changing Batch Operations CLI without\n  the CONFIRM gate. Job creation and priority changes can run up\n  significant costs and modify objects at scale.\n\n## Pre-flight safety checks (run before any remediation CLI)\n\nAll eight pre-flight safety checks (confirmation gate, cost estimation, manifest validation, role trust policy, KMS key state, Lambda state, report bucket region, rate-control sanity): [references\u002Fadvanced-patterns.md](references\u002Fadvanced-patterns.md).\n\n## Expert heuristic: the silent-completion trap\n\nThe silent-completion trap in full (the rule, why it exists, verification-technique table, three-step verification protocol, FAILURE_COUNT\u002FSAMPLE_VERIFIED surfacing, post-deploy detection): [references\u002Fadvanced-patterns.md](references\u002Fadvanced-patterns.md).\n\n## Recent AWS features (2024-2026)\n\nRecent AWS features detail (S3 Tables targets, server-side copy re-encrypt, FailedTasksOnly GA, per-job rate control, Glacier IR\u002FDeep Archive tiers, Object Lock operations, cross-region jobs, AWS\u002FS3Operations metrics): [references\u002Fadvanced-patterns.md](references\u002Fadvanced-patterns.md).\n\n## References (load on demand)\n\n- [Advanced patterns](references\u002Fadvanced-patterns.md) — Step 0 non-obvious Batch Operations behaviors, pre-flight safety checks, the silent-completion trap, recent AWS features\n- [Worked examples](references\u002Fworked-examples.md) — diagnose-job BLOCKED and Glacier bulk-restore COMPLETED walkthroughs\n- [Error handling](references\u002Ferror-handling.md) — job failure-mode table (`describe-job` symptom → root cause → fix)\n- [Diagnostic commands](references\u002Fdiagnostic-commands.md) — the eight live-account pre-flight probes\n\n## Domain\n\nAWS CloudOps \u002F S3 Batch Operations, Bulk Object Transformation, and\nAt-Scale Storage Management.\n\n## AWS documentation\n\n- **S3 Batch Operations User Guide** — https:\u002F\u002Fdocs.aws.amazon.com\u002FAmazonS3\u002Flatest\u002Fuserguide\u002Fbatch-ops.html\n- **Batch Operations operations** — https:\u002F\u002Fdocs.aws.amazon.com\u002FAmazonS3\u002Flatest\u002Fuserguide\u002Fbatch-ops-operations.html\n- **Creating a Batch Operations job** — https:\u002F\u002Fdocs.aws.amazon.com\u002FAmazonS3\u002Flatest\u002Fuserguide\u002Fbatch-ops-create-job.html\n- **Batch Operations IAM role** — https:\u002F\u002Fdocs.aws.amazon.com\u002FAmazonS3\u002Flatest\u002Fuserguide\u002Fbatch-ops-iam-role.html\n- **S3 inventory configuration** — https:\u002F\u002Fdocs.aws.amazon.com\u002FAmazonS3\u002Flatest\u002Fuserguide\u002Fstorage-inventory.html\n- **S3 Tables (Apache Iceberg)** — https:\u002F\u002Fdocs.aws.amazon.com\u002FAmazonS3\u002Flatest\u002Fuserguide\u002Fs3-tables.html\n- **S3 Control API Reference** — https:\u002F\u002Fdocs.aws.amazon.com\u002FAmazonS3\u002Flatest\u002FAPIReference\u002Fcontrol.html\n- **S3 Batch Operations pricing** — https:\u002F\u002Faws.amazon.com\u002Fs3\u002Fpricing\u002F\n","# S3 Batch Operations Operator\n\n## What this skill does\n\nExecutes S3 Batch Operations jobs correctly and safely at any scale. Runs\ndeterministic pre-checks before any state-changing CLI (manifest format\nand readability, report bucket writability, IAM role permission chain\nincluding operation-specific grants, KMS decrypt\u002Fencrypt path for\nSSE-KMS objects, Lambda invoke permission for Lambda operations,\nrate-control validity, manifest object-count estimate), executes the\n`create-job` behind a CONFIRM gate, and verifies the result by polling\n`describe-job` until `Status: Complete`, reading the completion report,\nand reconciling `Failed` count against an acceptable threshold. Every\njob plan names the operation-specific IAM grants; every diagnose-job\nsurfaces the failure-mode table so the operator knows whether the root\ncause is permissions, manifest format, throttling, or a per-object error.\n\n## Quick navigation\n\n| Section | What it covers | When to read it |\n|---|---|---|\n| **§ Quick reference** | Verdict thresholds (BLOCKED\u002FREADY\u002FCOMPLETED) + pre-check priority | Before any operation |\n| **§ Mindset** | Why a batch job is a manifest + role + report + rate-control quad, the false-green trap | Understanding the safety model |\n| **§ Pre-flight** | Job spec gate — manifest versioning, report bucket region, role ARN | Before executing any CLI |\n| **§ Process** | Per-operation planning: create-job, diagnose-job, cancel-job, update-priority | When choosing which operation to run |\n| **§ Output format** | STRICT output contract — OPERATION\u002FVERDICT\u002FPRE_CHECKS\u002FSTEPS\u002FPOST_VERIFY\u002FNOTES template | Formatting the response |\n| **§ Anti-Patterns** | NEVER list — common mistakes that silently fail jobs or skip objects | Review before risky operations |\n| **§ Pre-flight safety** | Capture pre-state, IAM role policy, KMS key policy, manifest validation | Defense-in-depth |\n| **§ Expert heuristic** | Batch Operations is async and per-object failures are silent — never trust `Active` alone | Avoiding the false-green trap |\n\n## STRICT output contract\n\nEVERY response MUST end with a single fenced text block in this exact\nshape (the operator's downstream tooling greps for it). No deviations:\n\n```text\nOPERATION: \u003Ccreate-job | diagnose-job | cancel-job | update-priority>\nVERDICT: READY | BLOCKED | COMPLETED\nTARGET: \u003Cmanifest-bucket\u002Fpath> (operation: \u003Ctype>, role: \u003Carn-or-\"none\">)\nPRE_CHECKS:\n  - [PASS] \u003Ccheck description>\n  - [FAIL] \u003Ccheck description> — \u003Creason>\nSTEPS:\n  1. \u003CCLI command with flags populated, or \"(none — pre-checks failed)\">\n  2. \u003Cwait \u002F monitoring command>\nPOST_VERIFY:\n  - [PASS] \u003Cverification description> | (pending execution)\n  - [FAIL] \u003Cverification description> — \u003Creason>\nNOTES: \u003Crate-control, monitoring, caveats>\n```\n\nRules of the contract:\n\n- VERDICT is exactly one of `READY`, `BLOCKED`, `COMPLETED`. No other\n  values. `READY` means pre-checks passed and a CONFIRM gate is\n  pending; `BLOCKED` means at least one pre-check failed — do NOT emit\n  STEPS that mutate state; `COMPLETED` means post-verification passed\n  after the job finished.\n- If PRE_CHECKS has any `[FAIL]`, VERDICT MUST be `BLOCKED` and STEPS\n  MUST be `(none — pre-checks failed)`.\n- If VERDICT is `READY`, STEPS[1] MUST be the CONFIRM prompt and\n  STEPS[2] MUST be the actual `create-job` CLI.\n- If VERDICT is `COMPLETED`, POST_VERIFY MUST contain at least one\n  `[PASS]` and no `[FAIL]`.\n- Never wrap the block in JSON, never abbreviate the field names,\n  never omit a section. If a section is empty, write `(none)`.\n\n## Quick reference — verdict thresholds\n\n| Verdict | Trigger condition | Action |\n|---|---|---|\n| `BLOCKED` | One or more pre-checks failed (manifest unreadable, report bucket not writable, role missing `batchoperations:*` or operation-specific grants, KMS decrypt denied for SSE-KMS source, Lambda resource policy missing `batchoperations.amazonaws.com` principal, rate-control invalid, manifest contains zero objects, destination bucket missing for copy) | List failures, do NOT execute |\n| `READY` | All pre-checks passed; awaiting CONFIRM gate | Emit exact `create-job` CLI with all flags populated, wait for operator yes |\n| `COMPLETED` | Job reached `Status: Complete` AND post-verification passed (`Failed` count within threshold, completion report readable, sample objects verified) | Emit verification results, monitoring plan |\n\n**Priority order for pre-checks (apply in this sequence, all must pass\nfor READY):**\n\n1. **Manifest format and readability.** Manifest is a versioned S3\n   inventory report (with `manifest.json` + `manifest.checksum`) OR a\n   CSV file (`bucket,key` per line, versioned CSV requires a third\n   column). The role can read the manifest object(s).\n2. **Report bucket configured and writable.** `ReportBucket` is in the\n   same Region as the Batch Operations job. Role has `s3:PutObject` on\n   the report prefix.\n3. **IAM role permission chain.** Role has `batchoperations:*` on the\n   job ARN (or `*`), plus read on the manifest bucket and write on the\n   target bucket(s). Operation-specific grants (see table below).\n4. **KMS decrypt path on source.** For SSE-KMS objects, the role has\n   `kms:Decrypt` on the source key ARN AND the source key policy\n   grants the role.\n5. **KMS encrypt path on destination.** For copy or re-encrypt\n   operations, the role has `kms:Encrypt` on the destination key ARN\n   AND the destination key policy grants the role.\n6. **Lambda invoke permission.** For invoke operations, the Lambda's\n   resource-based policy allows `batchoperations.amazonaws.com` to\n   `lambda:InvokeFunction`, scoped to the job role ARN.\n7. **Rate-control validity.** `RequestsPerSecond` (if set) is in\n   [1, 2500]. `CompletionWindow` (if set) is between 1 hour and 1 year.\n8. **Object-count estimate.** Manifest object count produces an ETA\n   within the configured `CompletionWindow`. For billion+ object\n   manifests, verify the rate limit produces a sub-window ETA.\n\n**Cost\u002Ftime baselines (2026):**\n\n- Per-object processing time: 0.5-2 seconds typical (operation-\n  dependent; Glacier restore and Lambda invoke are slower).\n- Default rate: S3 Batch Operations auto-scales within the account's\n  S3 request budget; explicit `RequestsPerSecond` caps it.\n- Job setup latency: 1-5 minutes from `create-job` to `Active`.\n- Completion report: written within minutes of job `Complete`.\n- Cost: per-request charges (PUT\u002FGET\u002Frestore) plus a Batch Operations\n  per-object processing fee (see S3 pricing).\n\n## Mindset\n\n**One-line takeaway:** `Status: Active` on a Batch Operations job is a\nscheduling claim, not a delivery claim. A job is only \"done\" when\n`Status: Complete`, the completion report shows the expected success\ncount, and a sample of objects verifies the operation actually applied.\nDriven by three S3 Batch Operations realities:\n\n- **A batch job is a manifest + role + report + rate-control quad.**\n  The manifest enumerates objects; the role grants the operation;\n  the report records results; the rate-control prevents runaway costs.\n  A single broken link makes the entire job fail or silently skip —\n  the job shows `Active` on dashboards while objects are untouched.\n- **Per-object failures are silent.** A job can complete with tens of\n  thousands of `Failed` entries while `Status: Complete` looks green.\n  Operators who do not read the completion report discover the failure\n  weeks later when downstream effects surface.\n- **Large jobs (billion+ objects) need explicit rate-control.** Without\n  `RequestsPerSecond`, the job can saturate the source or destination\n  bucket's request budget, crowd out application traffic, and exhaust\n  KMS request quotas (especially for decrypt+re-encrypt). Always set a\n  rate limit for jobs over 100 million objects.\n\n## Pre-flight: job spec gate\n\nRun before classification. Misclassifying these produces wrong plans.\n\n**Pagination:** `list-bucket-inventory-configurations` paginates at\n100. `describe-job` returns the full job descriptor in one call. CSV\nmanifests have no native pagination — estimate the object count from\nthe file size (each line averages 80-120 bytes for `bucket,key`).\n\n**Live-account pre-flight (skip if offline plan audit):**\nThe eight live-account pre-flight probes (manifest head-object, report-bucket region, versioning capture, role policy chain, KMS key state, destination KMS, Lambda policy, describe-job): [references\u002Fdiagnostic-commands.md](references\u002Fdiagnostic-commands.md).\n\n**Malformed input:** if the input JSON is invalid or missing required\nfields, emit `VERDICT: ERROR` with `REASON: Job spec is not valid JSON\nor is missing required fields — cannot plan.` and `REMEDIATION:\nRe-validate the spec against the operation schema in references\u002F\noperation-iam-matrix.md and re-plan.`\n\n| Job spec attribute | Effect on operation |\n|---|---|\n| `Manifest.Spec.Format: S3InventoryReport` | Requires `manifest.json` + `manifest.checksum`. Versioned inventory required for version-aware operations. |\n| `Manifest.Spec.Format: CSV` | Each line is `bucket,key` (unversioned) or `bucket,key,versionId` (versioned). No header row. |\n| `Operation` mismatch with manifest versioning | A versioned operation (e.g., object lock retention) on an unversioned manifest BLOCKS. |\n| `Report.Bucket` not in job Region | BLOCKED. Batch Operations requires the report bucket in the same Region as the job. |\n| `Role` missing `iam:PassRole` on the caller side | `create-job` itself fails with `AccessDenied`. Caller needs `iam:PassRole` on the role ARN. |\n| `RequestsPerSecond` outside [1, 2500] | BLOCKED. |\n| `Priority` outside [-2147483648, 2147483647] | BLOCKED. Higher value = higher priority. |\n| Manifest containing zero objects | BLOCKED. |\n| `ToggleEnabled: false` | Job created in `Suspended` state; will not run until `update-job-status` sets `Ready`. |\n\n## Process — operation planning (apply in order)\n\n### Step 0: Expert knowledge — non-obvious Batch Operations behaviors\n\nThese behaviors are easy to misjudge without operational experience.\nEach changes a plan if ignored:\nThe 12 non-obvious behaviors (two-sided role, versioned-inventory requirement, no CSV header row, fixed Lambda payload, two-stage Glacier restore, ServerSideCopy re-encrypt, full-tagset replace, global priority, CompletionWindow hint, report scope, S3 Tables operations, transient throttling): [references\u002Fadvanced-patterns.md](references\u002Fadvanced-patterns.md).\n\n\n### Step 1: Pre-check gate — BLOCKED if any check fails\n\nRun ALL pre-checks for the chosen operation. If ANY fails, the verdict\nis BLOCKED with the failed checks in PRE_CHECKS. Do NOT execute.\n\n**For ALL operations:**\n1. Manifest exists and is readable by the role.\n2. Manifest format matches the operation's versioning requirement.\n3. Report bucket exists in the same Region as the planned job.\n4. Caller has `s3control:CreateJob` AND `iam:PassRole` on the role.\n5. Role has `batchoperations:*` (or scoped equivalent).\n6. Role has read on the manifest bucket (`s3:GetObject` on the\n   manifest prefix).\n7. Role has `s3:PutObject` on the report bucket prefix.\n8. `RequestsPerSecond` (if set) is in [1, 2500].\n9. `Priority` is within the int32 range.\n10. Manifest object count is non-zero.\n\n**Operation-specific pre-checks:**\n\n| Operation | Additional checks |\n|---|---|\n| **Copy** | Destination bucket exists; role has `s3:PutObject` on destination; for SSE-KMS source, role has `kms:Decrypt`; for SSE-KMS destination, role has `kms:Encrypt` + destination key policy grants role. |\n| **ReplaceTag** | Role has `s3:PutObjectTagging` and `s3:DeleteObjectTagging` on the target. New tagset does not exceed 10 keys \u002F 256 chars per value. |\n| **Restore** | Source objects are in GLACIER \u002F GLACIER_IR \u002F DEEP_ARCHIVE. Role has `s3:RestoreObject`. `Days` and `GlacierJobTier` (Standard\u002FBulk) are set. |\n| **Replicate** | Source bucket has replication configured; destination rule exists. Role has `s3:ReplicateObject`, `s3:ReplicateDelete`. |\n| **Invoke** | Lambda `State: Active`. Lambda resource-based policy allows `batchoperations.amazonaws.com` to `lambda:InvokeFunction`. Lambda reserved concurrency >= `RequestsPerSecond`. Lambda timeout >= 60s (Batch retries on timeout, but a 3s timeout is almost always wrong). |\n| **PutACL** | Role has `s3:PutObjectAcl` on the target. Canned ACL or access-control policy is valid. |\n| **PutObjectLockRetention** | Target bucket has Object Lock enabled. Role has `s3:PutObjectRetention`. `Mode` is `GOVERNANCE` or `COMPLIANCE`. `RetainUntilDate` is in the future. |\n| **PutObjectLockLegalHold** | Target bucket has Object Lock enabled. Role has `s3:PutObjectLegalHold`. `Status` is `ON` or `OFF`. |\n\n**Job failure-mode table (use during diagnose-job):**\nThe nine-row failure-mode table mapping `describe-job` symptoms and completion-report failure codes to root causes and fixes: [references\u002Ferror-handling.md](references\u002Ferror-handling.md).\n\n### Step 2: READY — emit operation plan\n\nIf all pre-checks pass, emit `VERDICT: READY` with the exact CLI\nsequence and the CONFIRM gate. The plan includes:\n\n- The exact `aws s3control create-job` CLI with the JSON operation\n  spec inline or referenced from a file.\n- The expected duration (based on manifest size and rate limit).\n- The expected side-effects (object count processed, KMS request\n  volume, Lambda invocation count).\n- The CONFIRM gate prompt.\n- The monitoring commands (`describe-job` polling cadence, CloudWatch\n  metrics to watch).\n\n### Step 3: Execute behind CONFIRM gate\n\n- **MANDATORY CONFIRMATION GATE.** Before `create-job`,\n  `update-job-status`, `update-job-priority`, or any state-changing\n  CLI, emit: `CONFIRM: About to \u003Coperation> on job \u003Cid-or-\"new\"> in\n  account \u003Caccount> region \u003Cregion>. This will \u003Cconsequence>.\n  Proceed? (yes\u002Fno)`. Do NOT execute until the operator confirms.\n- Capture pre-state: if modifying an existing job,\n  `aws s3control describe-job --account-id \u003Caccount> --job-id \u003Cid>\n  --output json > \u002Ftmp\u002F\u003Cid>-pre-$(date +%s).json`.\n- For new jobs: write the operation spec to a local JSON file and\n  pass via `--operation`. Capture the returned `JobId`.\n- Poll `describe-job` every 60 seconds (or longer for billion-object\n  jobs) until `Status` transitions away from `Active`.\n\n### Step 4: Post-verification — COMPLETED\n\nAfter the job finishes, run post-verification. ALL checks must pass\nfor `COMPLETED`.\n\n1. `describe-job --job-id \u003Cid>` — confirm `Status: Complete`.\n2. Read `ProgressSummary` — `NumberOfTasksSucceeded` matches the\n   expected count; `NumberOfTasksFailed` is within the acceptable\n   threshold (typically 0; for billion-object jobs, \u003C 0.01%).\n3. Read the completion report (`Report.Bucket`). For `FailedTasksOnly`\n   scope, confirm the report is non-empty only if there were failures.\n4. Sample 5 objects from the manifest — verify the operation applied\n   (e.g., for copy, `head-object` on destination; for KMS re-encrypt,\n   verify `SSEKMSKeyId` matches the new key; for tag replace, verify\n   `get-object-tagging` matches the new tagset).\n5. For Glacier restore: sample 5 objects — verify `Restore` header\n   shows `ongoing-request=\"false\"` (or, for bulk tier, that the\n   request is queued).\n6. For Lambda invoke: check CloudWatch Logs for the Lambda — no\n   errors above the expected threshold.\n\nIf ANY verification fails, emit `VERDICT: ERROR` with the failure\ndetails — do not claim COMPLETED. A failed verification typically means\nthe job completed with unexpected failures; route to diagnose-job.\n\n## Output format (per operation)\n\nSee § STRICT output contract above. The block MUST appear exactly in\nthat shape at the end of the response.\n\n### Worked example — create-job (bulk KMS re-encrypt)\n\n```text\nOPERATION: create-job\nVERDICT: READY\nTARGET: s3:\u002F\u002Fprod-inventory\u002F2026-08-01\u002Fmanifest.json\n        (operation: copy with new SSE-KMS key, role:\n         arn:aws:iam::111111111111:role\u002FS3BatchOpsRole)\nPRE_CHECKS:\n  - [PASS] Manifest readable, format S3InventoryReport V2 (versioned)\n  - [PASS] Report bucket prod-batch-reports in us-east-1 (matches job)\n  - [PASS] Caller has s3control:CreateJob + iam:PassRole on\n    arn:aws:iam::111111111111:role\u002FS3BatchOpsRole\n  - [PASS] Role has batchoperations:*, s3:GetObject on prod-inventory,\n    s3:PutObject on prod-batch-reports\n  - [PASS] Role has s3:GetObject + s3:PutObject on prod-data\n  - [PASS] Role has kms:Decrypt on arn:aws:kms:us-east-1:111111111111:key\u002Fold-key\n  - [PASS] Role has kms:Encrypt on arn:aws:kms:us-east-1:111111111111:key\u002Fnew-key\n  - [PASS] New KMS key policy grants role Encrypt\n  - [PASS] RequestsPerSecond: 500 (in [1, 2500])\n  - [PASS] Manifest object count: 12,400,000 — ETA ~6.9h at 500 RPS\n    (within CompletionWindow: 24h)\nSTEPS:\n  1. CONFIRM: About to create a Batch Operations copy+re-encrypt job\n     on 12,400,000 objects in bucket prod-data (account\n     111111111111, region us-east-1), overwriting in place with new\n     SSE-KMS key arn:aws:kms:us-east-1:111111111111:key\u002Fnew-key, at\n     500 RPS. Estimated cost: ~$248 (PUT+GET+SSE-KMS). Proceed?\n     (yes\u002Fno)\n  2. aws s3control create-job \\\n       --account-id 111111111111 \\\n       --region us-east-1 \\\n       --priority 50 \\\n       --role-arn arn:aws:iam::111111111111:role\u002FS3BatchOpsRole \\\n       --operation '{\"S3InitiateRestoreObject\": null}' \\\n       --manifest '{\"Spec\":{\"Format\":\"S3InventoryReport\",\"Fields\":[\"Bucket\",\"Key\",\"VersionId\"]},\"Location\":{\"ObjectArn\":\"arn:aws:s3:::prod-inventory\u002F2026-08-01\u002Fmanifest.json\",\"ETag\":\"\u003Cetag>\"}}' \\\n       --report '{\"Bucket\":\"arn:aws:s3:::prod-batch-reports\",\"Prefix\":\"kms-reencrypt-2026-08\u002F\",\"Format\":\"Report_CSV_20180820\",\"ReportScope\":\"FailedTasksOnly\"}' \\\n       --client-request-token \"$(uuidgen)\"\n     (replace --operation with the actual copy-with-new-encryption JSON)\n  3. Poll: aws s3control describe-job --account-id 111111111111\n     --job-id \u003Creturned-id> --query 'Status'\n     (every 5 minutes for a 7-hour job)\nPOST_VERIFY:\n  - (pending execution)\nNOTES:\n  - Bulk KMS re-encrypt is a server-side copy with new SSE-KMS\n    metadata, overwriting in place. Set NewObjectMetadata.SSEKMSKeyId\n    to the new key in the operation JSON. Source ~12.4M Decrypt +\n    destination ~12.4M Encrypt requests. Verify KMS quota first.\n  - Use FailedTasksOnly report scope to avoid a multi-GB report.\n  - CloudWatch alarm on KMS Throttles during the job window.\n```\n\n### Worked example — diagnose-job (BLOCKED with remediation)\n\nFull diagnose-job BLOCKED example (Lambda reserved-concurrency root cause, remediation options, re-run of failed objects): [references\u002Fworked-examples.md](references\u002Fworked-examples.md).\n\n### Worked example — create-job (Glacier bulk restore, COMPLETED)\n\nFull Glacier bulk-restore COMPLETED example (two-stage restore verification): [references\u002Fworked-examples.md](references\u002Fworked-examples.md).\n\n## Anti-Patterns — NEVER\n\n- NEVER create a Batch Operations job without verifying the IAM role\n  has the operation-specific grants. `batchoperations:*` is the job-\n  control plane permission; the actual object operations need\n  `s3:GetObject`, `s3:PutObject`, `s3:RestoreObject`,\n  `s3:PutObjectTagging`, `s3:PutObjectRetention`, or\n  `s3:PutObjectAcl` depending on the operation type. Missing grants\n  produce a job that completes with 100% failure.\n\n- NEVER include a header row in a CSV manifest. The first line is\n  parsed as an object. A header `bucket,key` produces a `NoSuchKey`\n  failure for object `key` in bucket `bucket`. Use S3 inventory\n  format if you want structured metadata.\n\n- NEVER assume `Status: Complete` means all objects were successfully\n  processed. Read the completion report and reconcile\n  `NumberOfTasksFailed` against an acceptable threshold. For large\n  jobs, use `FailedTasksOnly` report scope and treat any non-empty\n  report as a follow-up item.\n\n- NEVER run a billion-object job without an explicit\n  `RequestsPerSecond`. Default auto-scaling can saturate the source\n  bucket, KMS quota, or Lambda concurrency, producing throttling\n  cascades that exhaust the retry budget.\n\n- NEVER assume Glacier restore is finished when the Batch Operations\n  job is `Complete`. The job submits restore requests; the actual\n  restore is asynchronous on the Glacier side. Poll `head-object\n  --restore` for `ongoing-request=\"false\"` before downstream use.\n\n- NEVER pass a role ARN at `create-job` without verifying the caller\n  has `iam:PassRole` on that ARN. The `create-job` API itself fails\n  with `AccessDenied` before any object processing.\n\n- NEVER set all jobs to `Priority: 2147483647`. Priority is a relative\n  scheduling signal within the account + Region. Inflating every job\n  to max nullifies the priority system and produces contention.\n\n- NEVER use the `Task` report scope for billion-object jobs. The\n  completion report will be a multi-GB CSV. Use `FailedTasksOnly` and\n  read the `ProgressSummary` for the success count.\n\n- NEVER attempt a `PutObjectLockRetention` operation on a bucket\n  without Object Lock enabled. The operation fails per-object with\n  `InvalidRequest`. Verify `get-object-lock-configuration` before\n  creating the job.\n\n- NEVER assume Lambda invocations are idempotent. Batch Operations\n  retries with the same `taskId`. If the Lambda has side effects\n  (writes to a database, sends a notification), it MUST deduplicate\n  on `taskId`.\n\n- NEVER trust `describe-job.ProgressSummary` alone for cost\n  verification. Cross-check S3 and KMS CloudWatch request metrics\n  during the job window — throttles and retries inflate actual\n  request count above the object count.\n\n- NEVER forget the manifest's `ETag` field. For S3 inventory\n  manifests, the `manifest.json` object has an ETag that must be\n  passed in the `Manifest.Location.ETag` field. Omitting it produces\n  a `400 MalformedXML` error from `create-job`.\n\n- NEVER auto-execute a state-changing Batch Operations CLI without\n  the CONFIRM gate. Job creation and priority changes can run up\n  significant costs and modify objects at scale.\n\n## Pre-flight safety checks (run before any remediation CLI)\n\nAll eight pre-flight safety checks (confirmation gate, cost estimation, manifest validation, role trust policy, KMS key state, Lambda state, report bucket region, rate-control sanity): [references\u002Fadvanced-patterns.md](references\u002Fadvanced-patterns.md).\n\n## Expert heuristic: the silent-completion trap\n\nThe silent-completion trap in full (the rule, why it exists, verification-technique table, three-step verification protocol, FAILURE_COUNT\u002FSAMPLE_VERIFIED surfacing, post-deploy detection): [references\u002Fadvanced-patterns.md](references\u002Fadvanced-patterns.md).\n\n## Recent AWS features (2024-2026)\n\nRecent AWS features detail (S3 Tables targets, server-side copy re-encrypt, FailedTasksOnly GA, per-job rate control, Glacier IR\u002FDeep Archive tiers, Object Lock operations, cross-region jobs, AWS\u002FS3Operations metrics): [references\u002Fadvanced-patterns.md](references\u002Fadvanced-patterns.md).\n\n## References (load on demand)\n\n- [Advanced patterns](references\u002Fadvanced-patterns.md) — Step 0 non-obvious Batch Operations behaviors, pre-flight safety checks, the silent-completion trap, recent AWS features\n- [Worked examples](references\u002Fworked-examples.md) — diagnose-job BLOCKED and Glacier bulk-restore COMPLETED walkthroughs\n- [Error handling](references\u002Ferror-handling.md) — job failure-mode table (`describe-job` symptom → root cause → fix)\n- [Diagnostic commands](references\u002Fdiagnostic-commands.md) — the eight live-account pre-flight probes\n\n## Domain\n\nAWS CloudOps \u002F S3 Batch Operations, Bulk Object Transformation, and\nAt-Scale Storage Management.\n\n## AWS documentation\n\n- **S3 Batch Operations User Guide** — https:\u002F\u002Fdocs.aws.amazon.com\u002FAmazonS3\u002Flatest\u002Fuserguide\u002Fbatch-ops.html\n- **Batch Operations operations** — https:\u002F\u002Fdocs.aws.amazon.com\u002FAmazonS3\u002Flatest\u002Fuserguide\u002Fbatch-ops-operations.html\n- **Creating a Batch Operations job** — https:\u002F\u002Fdocs.aws.amazon.com\u002FAmazonS3\u002Flatest\u002Fuserguide\u002Fbatch-ops-create-job.html\n- **Batch Operations IAM role** — https:\u002F\u002Fdocs.aws.amazon.com\u002FAmazonS3\u002Flatest\u002Fuserguide\u002Fbatch-ops-iam-role.html\n- **S3 inventory configuration** — https:\u002F\u002Fdocs.aws.amazon.com\u002FAmazonS3\u002Flatest\u002Fuserguide\u002Fstorage-inventory.html\n- **S3 Tables (Apache Iceberg)** — https:\u002F\u002Fdocs.aws.amazon.com\u002FAmazonS3\u002Flatest\u002Fuserguide\u002Fs3-tables.html\n- **S3 Control API Reference** — https:\u002F\u002Fdocs.aws.amazon.com\u002FAmazonS3\u002Flatest\u002FAPIReference\u002Fcontrol.html\n- **S3 Batch Operations pricing** — https:\u002F\u002Faws.amazon.com\u002Fs3\u002Fpricing\u002F\n",{"name":5,"license":15,"metadata":24,"description":7,"compatibility":42},{"tags":25,"phase":26,"author":27,"domain":28,"family":29,"version":30,"keywords":31,"task_type":32,"complexity":33,"entry_point":34,"skill_class":35,"when_to_use":36,"requires_llm":37,"verdict_shape":38,"lifecycle_status":39,"invocation_schema":40,"supports_pipeline":37,"activation_triggers":41},"aws, s3, s3control, storage, batch, bulk, glacier, kms, lambda, object-lock, replication, operate","4","Jacky Chan — AWS Community Builder","aws-cloudops","Storage","0.1.0","S3 Batch Operations, create-job, manifest, S3 inventory report, CSV manifest, completion report, job priority, RequestsPerSecond, rate control, copy operation, replace tag, restore from Glacier, replicate operation, invoke Lambda, put ACL, object lock retention, object lock legal hold, bulk storage class transition, bulk ACL fix, bulk replication backfill, bulk KMS re-encrypt, S3 Tables, billion objects, batchoperations, s3:GetObject, s3:PutObject","operate","high","false","capability","Creating an S3 Batch Operations job (copy, replace tag, restore from Glacier, replicate, invoke Lambda, put ACL, put object lock retention, put object lock legal hold), diagnosing a failed or stalled job, cancelling a runaway job, adjusting job priority or rate control, backfilling replication across existing objects, bulk-transitioning storage class, bulk-rotating KMS keys via decrypt + re-encrypt, invoking a custom Lambda across a manifest of objects, or planning a billion- object batch run with safe rate limits.","true","READY | BLOCKED | COMPLETED","active","Input: either (a) a job specification (operation type, source bucket, manifest format and location, report bucket, IAM role ARN, operation- specific parameters, priority, rate control) for plan classification, OR (b) a JobId for live-account diagnose or cancel operations. Output: deterministic OPERATION \u002F VERDICT \u002F PRE_CHECKS \u002F STEPS \u002F POST_VERIFY \u002F NOTES block per job, where VERDICT is one of READY, BLOCKED, COMPLETED.","S3 Batch Operations, batch copy objects, bulk restore Glacier, bulk replace tags, batch invoke Lambda, bulk replication backfill, bulk KMS re-encrypt, bulk storage class transition, put object lock retention batch, batch operations manifest, S3 inventory manifest, CSV manifest, completion report, RequestsPerSecond batch, cancel batch job, update job priority, S3 Tables batch operations, billion objects batch","Agent runtime that reads SKILL.md (Claude Code, Cursor, Windsurf, Codex, Gemini). No AWS CLI required for offline plan classification. Live-account operations use aws s3control create-job, describe-job, update-job-priority, update-job-status, aws s3api list-bucket-inventory-configurations, get-bucket-location, get-bucket-versioning, aws lambda get-policy, aws iam list-attached-role-policies \u002F list-role-policies, aws kms describe-key \u002F get-key-policy, and aws logs filter-log-events (AWS CLI...","main","eeace51bd5dd1db2030729d83eeab988861553aa","skills\u002Fs3-batch-operations-operator",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":52,"prompt":62,"zipAvailable":63},[53,56,59],{"tool":54,"command":55},"npx","npx skills add jhkchan\u002Faws-agent-skills --skill s3-batch-operations-operator",{"tool":57,"command":58},"bunx","bunx skills add jhkchan\u002Faws-agent-skills --skill s3-batch-operations-operator",{"tool":60,"command":61},"pnpm","pnpm dlx skills add jhkchan\u002Faws-agent-skills --skill s3-batch-operations-operator","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Fjhkchan\u002Faws-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Fs3-batch-operations-operator\u002FSKILL.md",true,[65,71,75,79,83,87,91,95,99,103,107,111,115,119,123,127,131,135,139,143,147],{"path":66,"resourceType":67,"resourceTypeDesc":68,"sizeBytes":69,"sourceHtmlUrl":70},"eval\u002Ftest-cases.yaml","REFERENCE","参考文档",8604,"https:\u002F\u002Fgithub.com\u002Fjhkchan\u002Faws-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Fs3-batch-operations-operator\u002Feval\u002Ftest-cases.yaml",{"path":72,"resourceType":67,"resourceTypeDesc":68,"sizeBytes":73,"sourceHtmlUrl":74},"evals\u002Fbaselines\u002Fcreate-copy-job-inventory-manifest-ready.md",859,"https:\u002F\u002Fgithub.com\u002Fjhkchan\u002Faws-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Fs3-batch-operations-operator\u002Fevals\u002Fbaselines\u002Fcreate-copy-job-inventory-manifest-ready.md",{"path":76,"resourceType":67,"resourceTypeDesc":68,"sizeBytes":77,"sourceHtmlUrl":78},"evals\u002Fbaselines\u002Fglacier-bulk-restore-completed.md",600,"https:\u002F\u002Fgithub.com\u002Fjhkchan\u002Faws-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Fs3-batch-operations-operator\u002Fevals\u002Fbaselines\u002Fglacier-bulk-restore-completed.md",{"path":80,"resourceType":67,"resourceTypeDesc":68,"sizeBytes":81,"sourceHtmlUrl":82},"evals\u002Fbaselines\u002Finvoke-lambda-job-blocked-permissions.md",887,"https:\u002F\u002Fgithub.com\u002Fjhkchan\u002Faws-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Fs3-batch-operations-operator\u002Fevals\u002Fbaselines\u002Finvoke-lambda-job-blocked-permissions.md",{"path":84,"resourceType":67,"resourceTypeDesc":68,"sizeBytes":85,"sourceHtmlUrl":86},"evals\u002Fbaselines\u002Fkms-reencrypt-job-blocked-key-policy.md",946,"https:\u002F\u002Fgithub.com\u002Fjhkchan\u002Faws-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Fs3-batch-operations-operator\u002Fevals\u002Fbaselines\u002Fkms-reencrypt-job-blocked-key-policy.md",{"path":88,"resourceType":67,"resourceTypeDesc":68,"sizeBytes":89,"sourceHtmlUrl":90},"evals\u002Fbaselines\u002Flarge-job-billion-objects-ready.md",873,"https:\u002F\u002Fgithub.com\u002Fjhkchan\u002Faws-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Fs3-batch-operations-operator\u002Fevals\u002Fbaselines\u002Flarge-job-billion-objects-ready.md",{"path":92,"resourceType":67,"resourceTypeDesc":68,"sizeBytes":93,"sourceHtmlUrl":94},"evals\u002Feval_config.yaml",1898,"https:\u002F\u002Fgithub.com\u002Fjhkchan\u002Faws-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Fs3-batch-operations-operator\u002Fevals\u002Feval_config.yaml",{"path":96,"resourceType":67,"resourceTypeDesc":68,"sizeBytes":97,"sourceHtmlUrl":98},"evals\u002Fevals.json",4441,"https:\u002F\u002Fgithub.com\u002Fjhkchan\u002Faws-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Fs3-batch-operations-operator\u002Fevals\u002Fevals.json",{"path":100,"resourceType":67,"resourceTypeDesc":68,"sizeBytes":101,"sourceHtmlUrl":102},"evals\u002Fprompts\u002Fcreate-copy-job-inventory-manifest-ready.md",1636,"https:\u002F\u002Fgithub.com\u002Fjhkchan\u002Faws-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Fs3-batch-operations-operator\u002Fevals\u002Fprompts\u002Fcreate-copy-job-inventory-manifest-ready.md",{"path":104,"resourceType":67,"resourceTypeDesc":68,"sizeBytes":105,"sourceHtmlUrl":106},"evals\u002Fprompts\u002Fglacier-bulk-restore-completed.md",1160,"https:\u002F\u002Fgithub.com\u002Fjhkchan\u002Faws-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Fs3-batch-operations-operator\u002Fevals\u002Fprompts\u002Fglacier-bulk-restore-completed.md",{"path":108,"resourceType":67,"resourceTypeDesc":68,"sizeBytes":109,"sourceHtmlUrl":110},"evals\u002Fprompts\u002Finvoke-lambda-job-blocked-permissions.md",1431,"https:\u002F\u002Fgithub.com\u002Fjhkchan\u002Faws-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Fs3-batch-operations-operator\u002Fevals\u002Fprompts\u002Finvoke-lambda-job-blocked-permissions.md",{"path":112,"resourceType":67,"resourceTypeDesc":68,"sizeBytes":113,"sourceHtmlUrl":114},"evals\u002Fprompts\u002Fkms-reencrypt-job-blocked-key-policy.md",1549,"https:\u002F\u002Fgithub.com\u002Fjhkchan\u002Faws-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Fs3-batch-operations-operator\u002Fevals\u002Fprompts\u002Fkms-reencrypt-job-blocked-key-policy.md",{"path":116,"resourceType":67,"resourceTypeDesc":68,"sizeBytes":117,"sourceHtmlUrl":118},"evals\u002Fprompts\u002Flarge-job-billion-objects-ready.md",1513,"https:\u002F\u002Fgithub.com\u002Fjhkchan\u002Faws-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Fs3-batch-operations-operator\u002Fevals\u002Fprompts\u002Flarge-job-billion-objects-ready.md",{"path":120,"resourceType":67,"resourceTypeDesc":68,"sizeBytes":121,"sourceHtmlUrl":122},"examples\u002FREADME.md",8455,"https:\u002F\u002Fgithub.com\u002Fjhkchan\u002Faws-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Fs3-batch-operations-operator\u002Fexamples\u002FREADME.md",{"path":124,"resourceType":67,"resourceTypeDesc":68,"sizeBytes":125,"sourceHtmlUrl":126},"references\u002Fadvanced-patterns.md",10523,"https:\u002F\u002Fgithub.com\u002Fjhkchan\u002Faws-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Fs3-batch-operations-operator\u002Freferences\u002Fadvanced-patterns.md",{"path":128,"resourceType":67,"resourceTypeDesc":68,"sizeBytes":129,"sourceHtmlUrl":130},"references\u002Fdiagnostic-commands.md",1475,"https:\u002F\u002Fgithub.com\u002Fjhkchan\u002Faws-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Fs3-batch-operations-operator\u002Freferences\u002Fdiagnostic-commands.md",{"path":132,"resourceType":67,"resourceTypeDesc":68,"sizeBytes":133,"sourceHtmlUrl":134},"references\u002Ferror-handling.md",1913,"https:\u002F\u002Fgithub.com\u002Fjhkchan\u002Faws-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Fs3-batch-operations-operator\u002Freferences\u002Ferror-handling.md",{"path":136,"resourceType":67,"resourceTypeDesc":68,"sizeBytes":137,"sourceHtmlUrl":138},"references\u002Flarge-job-playbook.md",6077,"https:\u002F\u002Fgithub.com\u002Fjhkchan\u002Faws-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Fs3-batch-operations-operator\u002Freferences\u002Flarge-job-playbook.md",{"path":140,"resourceType":67,"resourceTypeDesc":68,"sizeBytes":141,"sourceHtmlUrl":142},"references\u002Foperation-iam-matrix.md",7333,"https:\u002F\u002Fgithub.com\u002Fjhkchan\u002Faws-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Fs3-batch-operations-operator\u002Freferences\u002Foperation-iam-matrix.md",{"path":144,"resourceType":67,"resourceTypeDesc":68,"sizeBytes":145,"sourceHtmlUrl":146},"references\u002Fworked-examples.md",2679,"https:\u002F\u002Fgithub.com\u002Fjhkchan\u002Faws-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Fs3-batch-operations-operator\u002Freferences\u002Fworked-examples.md",{"path":148,"resourceType":67,"resourceTypeDesc":68,"sizeBytes":149,"sourceHtmlUrl":150},"skill.usf.yaml",4703,"https:\u002F\u002Fgithub.com\u002Fjhkchan\u002Faws-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Fs3-batch-operations-operator\u002Fskill.usf.yaml",[152,159,166,173,180,187],{"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},"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":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},"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":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},"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":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},"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":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},"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":188,"name":189,"nameZh":6,"description":190,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":191,"categories":192,"keywords":193,"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",[],[]]