[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M7AB4H2D40DWF9PFS1HF7934":3,"skill-same:01M7AB4H2D40DWF9PFS1HF7934":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},"01M7AB4H2D40DWF9PFS1HF7934","rds-upgrade-operator",null,"Operates Amazon RDS and Aurora engine upgrade workflows end-to-end — major version upgrades (explicit opt-in, pre-check: snapshot, parameter group compatibility, application driver support), minor version upgrades (auto-minor-version flag, patch schedule), Aurora MySQL 5.7 to 8.0, PostgreSQL 13 to 14 to 15 upgrade paths, blue\u002Fgreen deployments for zero- downtime upgrades, custom parameter group migration for the target engine version, option group migration, application connection string validation, post-upgrade verification (query performance regression, feature parity, analyzer changes), rollback strategy via point-in-time recovery, global database upgrade sequencing (primary then secondaries), multi-AZ upgrade behavior (rolling failover), and maintenance window scheduling. Runs deterministic pre-checks (available state, pending actions, snapshot exists, param group compatibility, option group compatibility, replication topology, storage type, free storage headroom) behind a CONFIRM gate and emits...","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\u002Frds-upgrade-operator\u002FSKILL.md",[],[],"2026-11-15 14:19:17",false,"---\nname: rds-upgrade-operator\ndescription: 'Operates Amazon RDS and Aurora engine upgrade workflows end-to-end — major version upgrades (explicit opt-in, pre-check: snapshot, parameter group compatibility, application driver support), minor version upgrades (auto-minor-version flag, patch schedule), Aurora MySQL 5.7 to 8.0, PostgreSQL 13 to 14 to 15 upgrade paths, blue\u002Fgreen deployments for zero- downtime upgrades, custom parameter group migration for the target engine version, option group migration, application connection string validation, post-upgrade verification (query performance regression, feature parity, analyzer changes), rollback strategy via point-in-time recovery, global database upgrade sequencing (primary then secondaries), multi-AZ upgrade behavior (rolling failover), and maintenance window scheduling. Runs deterministic pre-checks (available state, pending actions, snapshot exists, param group compatibility, option group compatibility, replication topology, storage type, free storage headroom) behind a CONFIRM gate and emits...'\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 rds modify-db-instance --engine-version, modify-db-cluster --engine-version, create-blue-green-deployment, switchover-blue-green-deployment, describe-db-engine-versions, describe-db-instances, describe-db-clusters, create-db-snapshot, describe-db-snapshots, modify-db-instance --auto-minor-version-upgrade, promote-read-replica...\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: Databases\n  task_type: operate\n  skill_class: capability\n  lifecycle_status: active\n  verdict_shape: OPERATION_COMPLETED | REVIEW_REQUIRED\n  when_to_use: Upgrading an RDS instance or Aurora cluster engine version (major or minor), planning a major version cut-over with pre-checks and rollback, deploying blue\u002Fgreen for zero-downtime upgrade, migrating parameter or option groups to a target engine version, sequencing a global database upgrade across Regions, scheduling a maintenance-window upgrade, enabling or disabling auto-minor-version-upgrade, validating application connection strings after an upgrade, or verifying post-upgrade query performance and feature parity.\n  activation_triggers: upgrade RDS engine, major version upgrade, minor version upgrade, Aurora MySQL 8.0 upgrade, PostgreSQL 15 upgrade, blue\u002Fgreen deploy RDS, zero-downtime database upgrade, parameter group migration, option group migration, global database upgrade, maintenance window upgrade, auto minor version upgrade, post-upgrade verification, rollback RDS upgrade, RDS engine version\n  invocation_schema: 'Input: either (a) a database\u002Fcluster configuration (describe-db-instances or describe-db-clusters output) plus the intended operation (major-upgrade, minor-upgrade, blue-green-upgrade, param-group-migrate, global-upgrade, configure-auto-minor, post-upgrade-verify, rollback), OR (b) a DB instance or cluster identifier + operation for live-account execution. Output: deterministic OPERATION \u002F VERDICT \u002F PRE_CHECKS \u002F STEPS \u002F POST_VERIFY \u002F NOTES block per upgrade, where VERDICT is OPERATION_COMPLETED or REVIEW_REQUIRED.'\n  version: 0.1.0\n  author: Jacky Chan — AWS Community Builder\n  keywords: RDS upgrade, Aurora upgrade, major version upgrade, minor version upgrade, engine version, auto-minor-version-upgrade, Aurora MySQL 8.0, PostgreSQL 15, blue\u002Fgreen deployment, zero-downtime upgrade, parameter group migration, option group migration, connection string validation, post-upgrade verification, query performance regression, point-in-time recovery, rollback strategy, global database upgrade, multi-AZ upgrade, maintenance window, switchover, UpgradeInProgress\n  tags: aws, rds, aurora, database, upgrade, mysql, postgresql, blue-green, operate\n---\n\n# RDS and Aurora Engine Upgrade Operator\n\n## What this skill does\n\nExecutes RDS and Aurora engine upgrade operations correctly and safely.\nRuns deterministic pre-checks before any state-changing CLI (instance or\ncluster is `available`, no pending maintenance actions, pre-upgrade\nsnapshot exists and is `available`, parameter group is compatible with the\ntarget engine version, option group is compatible, replication topology\nis understood, storage type and free-space headroom are sufficient),\nexecutes the upgrade behind a CONFIRM gate, and verifies the result by\nconfirming `EngineVersion` matches the target, the instance or cluster\nreturns to `available`, application connectivity is validated, and no\nquery-performance regression exceeds the configured threshold. Every\nmajor upgrade produces an explicit opt-in requirement plus a blue\u002Fgreen\nrecommendation for zero downtime; every minor upgrade surfaces the\n`AutoMinorVersionUpgrade` flag so the operator knows whether the next\npatch is automatic.\n\n## Quick navigation\n\n| Section | What it covers | When to read it |\n|---|---|---|\n| **§ Quick reference** | Verdict thresholds (REVIEW_REQUIRED\u002FOPERATION_COMPLETED) + pre-check priority | Before any operation |\n| **§ Mindset** | Why major upgrades need explicit opt-in, blue\u002Fgreen as the zero-downtime primitive, the param-group trap | Understanding the safety model |\n| **§ Pre-flight** | Database\u002Fcluster metadata gate — state, pending actions, multi-AZ, global cluster, replicas | Before executing any CLI |\n| **§ Process** | Per-operation planning: major, minor, blue-green, param-group, global, auto-minor, verify, rollback | When choosing which operation to run |\n| **§ Output format** | Structured OPERATION\u002FVERDICT\u002FPRE_CHECKS\u002FSTEPS\u002FPOST_VERIFY template | Formatting the response |\n| **§ Anti-Patterns** | NEVER list — common mistakes that cause extended downtime or strand a cluster mid-upgrade | Review before risky operations |\n| **§ Pre-flight safety** | Capture pre-state, snapshot, application driver check, rollback plan | Defense-in-depth |\n\n## Quick reference — verdict thresholds\n\n| Verdict | Trigger condition | Action |\n|---|---|---|\n| `REVIEW_REQUIRED` | Operation plan is ready but requires human review before execution (major upgrade requires explicit opt-in, blue\u002Fgreen plan needs cutover timing approval, global database upgrade sequencing across Regions, PITR rollback plan, parameter group with `pending-reboot` changes that need a maintenance window) | Emit plan with the specific items to review, wait for operator approval |\n| `OPERATION_COMPLETED` | Upgrade finished and post-verification passed (`EngineVersion` matches target, instance or cluster returned to `available`, application connectivity validated, no query regression above threshold, replicas in sync) | Emit verification results, monitoring plan |\n\n**Priority order for pre-checks (apply in this sequence, any failure routes\nto REVIEW_REQUIRED with the failure surfaced):**\n\n1. **Database reachability** — instance or cluster exists, `DBInstanceStatus`\n   or `Status` is `available`, NOT `modifying`, `upgrading`,\n   `backing-up`, or `storage-optimization`. No `PendingModifiedValues`\n   that would conflict with the upgrade.\n2. **Target engine version validity** — `describe-db-engine-versions`\n   confirms the target `EngineVersion` is valid for the instance class,\n   storage type, and current engine. Major version path is supported\n   (skipping major versions is not allowed).\n3. **Pre-upgrade snapshot** — a manual snapshot exists taken AFTER any\n   pending writes, with status `available`. This is the rollback anchor.\n4. **Parameter group compatibility** — a custom parameter group exists for\n   the target engine family (e.g., `default.aurora-mysql8.0` for an Aurora\n   MySQL 5.7 to 8.0 upgrade). The default param group from the source\n   version is NOT compatible.\n5. **Option group compatibility** — an option group for the target engine\n   family exists and includes the same options (or validated replacements).\n6. **Replication topology understood** — read replicas, Aurora replicas,\n   and global cluster secondaries are identified. The upgrade order is\n   primary then replicas (or blue\u002Fgreen switchover).\n7. **Storage headroom** — at least 20% free storage for the upgrade\n   temporary space (engine upgrade can rebuild system tables).\n8. **Application driver support** — the application's database driver\n   version supports the target engine version (e.g., MySQL Connector\u002FJ 8.x\n   for Aurora MySQL 8.0; PostgreSQL JDBC 42.x for PostgreSQL 15).\n9. **Maintenance window** — the upgrade is scheduled in the configured\n   maintenance window OR the operator has explicitly approved an immediate\n   upgrade.\n\n**Cost\u002Ftime baselines (2026):**\n\n- Minor version upgrade: 5-20 minutes of `upgrading` state, brief\n  connection drop on single-AZ; multi-AZ rolls through a failover.\n- Major version upgrade: 20 minutes to 4 hours depending on database\n  size (system table rebuild). Aurora is typically faster than RDS\n  Standard because Aurora stores the system catalogs differently.\n- Blue\u002Fgreen switchover: typically under 60 seconds (DNS shift). The\n  green environment provisioning takes 30-90 minutes before the\n  switchover.\n- Global database upgrade: primary Region first (20 min to 4 hours),\n  then each secondary Region sequentially (the secondary rebuilds from\n  the upgraded primary).\n- PITR rollback to pre-upgrade state: recovery time objective is the\n  snapshot-or-PITR-restore time (minutes to hours depending on size).\n\n## Mindset\n\n**One-line takeaway:** a major version upgrade is an irreversible\ncut-over unless you pre-planned a rollback path (PITR to the pre-upgrade\nmoment or a blue\u002Fgreen switchover back). `modify-db-instance\n--engine-version` returns immediately but the engine upgrade runs\nasynchronously — the database is in `upgrading` state for the full\nduration, during which the old version is gone and the new version is\nnot yet ready. Driven by three RDS realities:\n\n- **Major upgrades require explicit opt-in; auto applies only minor.**\n  `AutoMinorVersionUpgrade: true` schedules minor version patches in the\n  maintenance window. It NEVER triggers a major version upgrade. Setting\n  `--engine-version` to a higher major version is the explicit opt-in\n  and it is irreversible without a restore — you cannot \"downgrade\" an\n  engine in place.\n\n- **Blue\u002Fgreen deploy provides zero-downtime upgrade.** The blue\u002Fgreen\n  pattern creates a full copy of the production environment (the green)\n  running the target engine version, keeps it in sync via logical\n  replication, and switches traffic over via a DNS shift (typically\n  under 60 seconds). This is the recommended path for production major\n  upgrades. The switchover is reversible until the green is deleted.\n\n- **Parameter group must be compatible with the target version.** Each\n  engine family has its own parameter group family (e.g.,\n  `aurora-mysql5.7` vs `aurora-mysql8.0`). An upgrade from 5.7 to 8.0\n  REQUIRES a parameter group from the `aurora-mysql8.0` family. Using\n  the 5.7 parameter group causes the upgrade to fail or the instance to\n  boot with default parameters, silently dropping custom settings like\n  `max_connections`, `innodb_buffer_pool_size`, or `character_set_server`.\n\n## Pre-flight: database metadata gate\n\nRun before classification. Misclassifying these produces wrong plans.\n\n**Pagination:** `describe-db-instances` paginates at 100\u002Fpage — drain\n`--starting-token` to completion.\n\nSix live-account pre-flight commands (instance and cluster metadata, valid upgrade targets, pre-upgrade snapshots, pending maintenance, global topology): moved verbatim to [references\u002Fdiagnostic-commands.md](references\u002Fdiagnostic-commands.md).\nThe attribute table below maps the captured fields to plan effects.\n\n**Malformed input:** if the input JSON is invalid or missing required\nfields, emit `VERDICT: ERROR` with `REASON: DB\u002Fcluster configuration\nis not valid JSON or is missing required fields — cannot plan.` and\n`REMEDIATION: Re-fetch with aws rds describe-db-instances\n--db-instance-identifier \u003Cid> --output json and re-plan.`\n\n| Database attribute | Effect on operation |\n|---|---|\n| `DBInstanceStatus` not `available` | Upgrade BLOCKED. Wait for the current operation to finish, or investigate a stuck status. |\n| `PendingModifiedValues` non-empty | Upgrade routes to REVIEW_REQUIRED. Resolve the pending change (apply immediately or cancel) before scheduling the upgrade. |\n| `MultiAZ: true` | Upgrade runs through a rolling failover. Plan for two brief connection drops (primary -> standby, then upgrade -> failback). |\n| `ReadReplicaSourceDBInstanceIdentifier` set (this is a replica) | Upgrade the SOURCE first, then replicas. Replicas cannot run a higher engine version than the source. |\n| `GlobalClusterIdentifier` set | Global cluster member. Primary Region must upgrade first, then secondaries. Use the global-upgrade operation. |\n| `StorageType: gp2` with \u003C 20% free | Upgrade routes to REVIEW_REQUIRED — increase storage or convert to gp3 first. |\n| `AutoMinorVersionUpgrade: true` | Minor versions are applied automatically in the maintenance window. A manual minor upgrade overrides the schedule. |\n| `DBParameterGroups[0].ParameterApplyStatus: pending-reboot` | A pending parameter change will be applied during the upgrade reboot. Route to REVIEW_REQUIRED. |\n| `DBParameterGroups[0].DBParameterGroupName: default.\u003Cengine>\u003Cversion>` | Default param group. For major upgrades, a custom param group for the target version is required. |\n| `OptionGroupMemberships[0].Status: pending-maintenance` | An option group change is pending. Resolve before upgrade. |\n\n## Process — operation planning (apply in order)\n\n### Step 0: Expert knowledge — non-obvious RDS\u002FAurora upgrade behaviors\n\nTen non-obvious upgrade behaviors (explicit major opt-in, parameter-family migration, blue\u002Fgreen zero-downtime path, multi-AZ rolling failover, apply-immediately semantics, caching_sha2_password, PostgreSQL 14+ statistics, Serverless v1 restore-only, SupportsGlobalDatabases, upgrade-failed recovery): moved verbatim to [references\u002Fadvanced-patterns.md](references\u002Fadvanced-patterns.md).\nLoad before planning any upgrade; Steps 1-4 assume these constraints.\n\n### Step 1: Pre-check gate — REVIEW_REQUIRED if any check needs human attention\n\nRun ALL pre-checks for the chosen operation. If any check fails or\nrequires human review, the verdict is REVIEW_REQUIRED with the specific\nitems listed. Do NOT execute until the operator reviews.\n\n**For ALL operations:**\n1. Instance or cluster exists (`describe-db-instances` \u002F\n   `describe-db-clusters` does not return\n   `DBInstanceNotFound`).\n2. `DBInstanceStatus` \u002F `Status` is `available` (not `modifying`,\n   `upgrading`, `backing-up`, `storage-optimization`, `resetting`,\n   `incompatible-parameters`).\n3. No `PendingModifiedValues` (a pending change conflicts with the\n   upgrade — apply or cancel first).\n4. Storage type and free space: at least 20% free storage. For `gp2`,\n   consider converting to `gp3` first (gp3 supports higher IOPS at\n   lower cost and is the recommended storage type for 2024+).\n5. `PerformanceInsightsEnabled: true` (for before\u002Fafter comparison).\n\n**For major-upgrade (`modify-db-instance --engine-version\n\u003Ctarget-major>`):**\n6. `describe-db-engine-versions` confirms the target version is a\n   valid upgrade target from the current version.\n7. The upgrade path is sequential (no skipped major versions).\n8. A custom parameter group for the target engine family exists and\n   is attached (or will be attached as part of the modify call).\n9. An option group for the target engine family exists and is attached\n   (or will be attached as part of the modify call).\n10. A pre-upgrade manual snapshot exists and is `available`.\n11. Application driver support verified (driver version supports the\n    target engine version).\n12. `PreferredMaintenanceWindow` acknowledged — with\n    `--apply-immediately` the upgrade starts now; without it, the\n    upgrade is deferred to the next window.\n13. Blue\u002Fgreen recommended for production (surface the alternative\n    plan if the operator chose in-place).\n\n**For minor-upgrade (`modify-db-instance --engine-version\n\u003Ctarget-minor>` or `--auto-minor-version-upgrade`):**\n6. `describe-db-engine-versions` confirms the target version is valid.\n7. Minor version is within the same major version.\n8. `AutoMinorVersionUpgrade` current state acknowledged — if `true`\n   and the target minor is the current preferred patch, the upgrade\n   may happen automatically in the next window.\n\n**For blue-green-upgrade (`create-blue-green-deployment`):**\n6. The source instance or cluster is `available`.\n7. The target engine version is a valid upgrade target.\n8. Parameter and option groups for the target version are identified.\n9. The application tolerates logical replication lag during the sync\n   phase (typically seconds; longer for large write-heavy workloads).\n10. A switchover window is planned (the switchover itself is under 60\n    seconds but the green provisioning takes 30-90 minutes).\n\n**For param-group-migrate (pre-stage a target-version parameter group):**\n6. The target parameter group family matches the target engine family\n   (e.g., `aurora-mysql8.0` for Aurora MySQL 8.0).\n7. Parameters diffed against the source param group; custom values\n   reproduced.\n8. Static parameters (those requiring a reboot) identified — they will\n   apply at the next reboot (i.e., during the upgrade).\n\n**For global-upgrade (multi-Region global cluster):**\n6. `describe-global-clusters` confirms the global cluster topology.\n7. The primary Region is identified.\n8. Each secondary Region's cluster is `available`.\n9. The target engine version has `SupportsGlobalDatabases: true`.\n10. A Region-by-Region cutover plan is documented (primary first,\n    secondaries rebuilt sequentially).\n\n**For configure-auto-minor (toggle `AutoMinorVersionUpgrade`):**\n6. The operator understands that enabling auto-minor schedules patches\n   in the maintenance window — it does NOT trigger an immediate patch.\n\n**For post-upgrade-verify (read-only):**\n6. The instance or cluster has returned to `available`.\n7. `EngineVersion` matches the target.\n8. Performance Insights has data for the pre-upgrade and post-upgrade\n   windows.\n\n**For rollback (PITR to pre-upgrade state):**\n6. A pre-upgrade snapshot or a valid PITR window (within the backup\n   retention period) exists.\n7. The operator understands that rollback means a NEW instance or\n   cluster — the endpoint changes. Application connection strings\n   must be updated.\n\nUpgrade failure-mode table (InvalidParameterCombination, parameter\u002Foption group gaps, stuck upgrade-failed, MySQL 8.0 auth plugin, pg_hba.conf, planner-statistics regressions, replica lag, global secondary rebuild): moved verbatim to [references\u002Ferror-handling.md](references\u002Ferror-handling.md).\nLoad when a pre-check fails or an upgrade misbehaves.\n\n### Step 2: OPERATION_COMPLETED or REVIEW_REQUIRED — emit operation plan\n\nIf all pre-checks pass AND the operation is read-only or low-risk\n(minor version with `--apply-immediately`, parameter group migration\nwithout static changes, post-upgrade verification), emit\n`VERDICT: OPERATION_COMPLETED` (for already-finished operations) or\nthe executable plan with the CONFIRM gate.\n\nFor operations requiring human review (major upgrade, blue\u002Fgreen\ncutover, global database sequencing, rollback via PITR), emit\n`VERDICT: REVIEW_REQUIRED` with the specific items to review. The plan\nincludes:\n\n- The exact AWS CLI command with all flags populated from the database\n  configuration.\n- The expected duration (minor 5-20 min; major 20 min to 4 hours;\n  blue\u002Fgreen switchover under 60 seconds after 30-90 min of green\n  provisioning).\n- The expected side-effects (`EngineVersion` changes, instance reboots,\n  connections drop, replicas rebuild).\n- The CONFIRM gate prompt.\n- The rollback plan (snapshot ARN for in-place; blue\u002Fgreen switchback\n  for blue\u002Fgreen).\n\n### Step 3: Execute behind CONFIRM gate\n\n- **MANDATORY CONFIRMATION GATE.** Before any state-changing CLI\n  (`modify-db-instance`, `modify-db-cluster`,\n  `create-blue-green-deployment`, `switchover-blue-green-deployment`,\n  `delete-blue-green-deployment`, `promote-read-replica`,\n  `restore-db-instance-to-point-in-time`), emit:\n  `CONFIRM: About to \u003Coperation> on \u003Cdb-id> in account \u003Caccount>\n  region \u003Cregion>. This will \u003Cconsequence>. Proceed? (yes\u002Fno)`.\n  Do NOT execute until the operator confirms.\n- Capture pre-state for rollback: `aws rds describe-db-instances\n  --db-instance-identifier \u003Cid> --output json > \u002Ftmp\u002F\u003Cid>-pre-$(date\n  +%s).json` AND a manual snapshot if one does not already exist:\n  `aws rds create-db-snapshot --db-instance-identifier \u003Cid>\n  --db-snapshot-id \u003Cid>-pre-upgrade-$(date +%s)`.\n- Execute the CLI. For `modify-db-instance`, the API returns\n  immediately; the actual upgrade runs asynchronously. The instance\n  enters `upgrading` state.\n- Monitor the upgrade via RDS events:\n  `aws rds describe-events --source-type db-instance\n  --source-identifier \u003Cid> --start-time $(($(date +%s) - 600))000`.\n\n### Step 4: Post-verification — OPERATION_COMPLETED\n\nAfter the upgrade finishes, run post-verification. ALL checks must pass\nfor `OPERATION_COMPLETED`.\n\n1. `describe-db-instances --db-instance-identifier \u003Cid>` — confirm\n   `EngineVersion` matches the target AND `DBInstanceStatus: available`.\n2. `describe-db-snapshots` — confirm the automatic snapshot taken by RDS\n   during the upgrade is `available` (this is the post-upgrade backup).\n3. For multi-AZ: confirm both the primary and standby are on the target\n   version (check `SecondaryAvailabilityZone` and failover status).\n4. For Aurora clusters: confirm ALL cluster members (writer and readers)\n   report the target `EngineVersion`. Check `AuroraReplicaLag` has\n   returned to baseline.\n5. For global clusters: confirm ALL secondary Region clusters have been\n   rebuilt and report the target version.\n6. Application connectivity: test a connection from the application's\n   subnet\u002FSG to the database endpoint on the DB port. Verify the\n   application driver version connects successfully.\n7. Query performance verification: compare Performance Insights\n   top-queries latency (post-upgrade) against the pre-upgrade baseline.\n   Investigate any query with > 2x latency regression.\n8. For PostgreSQL: run `ANALYZE` on all user databases to refresh\n   planner statistics. Schedule a `VACUUM ANALYZE` during the next low-\n   traffic window.\n9. For MySQL\u002FAurora MySQL 8.0: verify the authentication plugin\n   (`caching_sha2_password` vs `mysql_native_password`) matches what\n   the application driver expects.\n\nIf ANY verification fails, emit `VERDICT: ERROR` with the failure\ndetails — do not claim OPERATION_COMPLETED. A failed verification\ntypically means a query regression or an application connectivity\nissue; route to the rollback operation if the regression is severe.\n\n## Output format (per operation)\n\n```text\nOPERATION: \u003Cmajor-upgrade | minor-upgrade | blue-green-upgrade | param-group-migrate | global-upgrade | configure-auto-minor | post-upgrade-verify | rollback>\nVERDICT: OPERATION_COMPLETED | REVIEW_REQUIRED\nTARGET: \u003Cdb-id-or-cluster-id> (engine: \u003Ccurrent-version> -> \u003Ctarget-version>)\nPRE_CHECKS:\n  - [PASS] \u003Ccheck description>\n  - [REVIEW] \u003Ccheck description> — \u003Citem to review>\nSTEPS:\n  1. \u003CCLI command with flags populated>\n  2. \u003Cwait \u002F monitoring command>\n  3. \u003Cnext step>\nPOST_VERIFY:\n  - [PASS] \u003Cverification description>\n  - [FAIL] \u003Cverification description> — \u003Creason>\nNOTES: \u003Crollback plan, monitoring, caveats>\n```\n\n### Worked example — minor-upgrade (in-place, OPERATION_COMPLETED)\n\n```text\nOPERATION: minor-upgrade\nVERDICT: OPERATION_COMPLETED\nTARGET: prod-orders-db (engine: postgres 14.10 -> 14.11)\nPRE_CHECKS:\n  - [PASS] DBInstanceStatus: available\n  - [PASS] No PendingModifiedValues\n  - [PASS] describe-db-engine-versions confirms 14.11 is valid for\n    aurora-postgresql on db.r6g.large\n  - [PASS] Storage: 800 GB allocated, 350 GB free (44% headroom)\n  - [PASS] PerformanceInsightsEnabled: true\n  - [PASS] AutoMinorVersionUpgrade: false (operator-initiated)\nSTEPS:\n  1. CONFIRM: About to upgrade prod-orders-db from PostgreSQL 14.10 to\n     14.11 in account 111111111111 region us-east-1. This will reboot\n     the instance (brief connection drop, ~30 seconds for single-AZ,\n     two drops for multi-AZ). Proceed? (yes\u002Fno)\n  2. aws rds create-db-snapshot \\\n       --db-instance-identifier prod-orders-db \\\n       --db-snapshot-id prod-orders-db-pre-14-11-$(date +%s)\n  3. aws rds wait db-snapshot-available \\\n       --db-snapshot-id prod-orders-db-pre-14-11-...\n  4. aws rds modify-db-instance \\\n       --db-instance-identifier prod-orders-db \\\n       --engine-version 14.11 \\\n       --apply-immediately\n  5. aws rds wait db-instance-available \\\n       --db-instance-identifier prod-orders-db\nPOST_VERIFY:\n  - [PASS] EngineVersion: 14.11 (confirmed via describe-db-instances)\n  - [PASS] DBInstanceStatus: available\n  - [PASS] Application connectivity: psql connect from app subnet OK\n  - [PASS] Performance Insights: top-10 query latency within +\u002F- 10%\n    of pre-upgrade baseline\nNOTES:\n  - Minor version upgrade completed in 12 minutes. Single-AZ, one\n    connection drop of ~25 seconds.\n  - PostgreSQL ANALYZE scheduled for the next low-traffic window\n    (03:00 UTC) to refresh planner statistics.\n  - Rollback: restore from snapshot prod-orders-db-pre-14-11-...\n    if any regression is detected. This creates a NEW instance —\n    endpoint will change.\n```\n\n### Worked example — major-upgrade via blue\u002Fgreen (REVIEW_REQUIRED)\n\nFull worked example (Aurora MySQL 5.7 to 8.0 blue\u002Fgreen with driver and switchover review items): moved verbatim to [references\u002Fworked-examples.md](references\u002Fworked-examples.md).\nApply the Output format template above to reproduce it.\n\n### Worked example — global database upgrade (condensed)\n\nCondensed worked example (global database upgrade with Region sequencing): moved verbatim to [references\u002Fworked-examples.md](references\u002Fworked-examples.md).\nApply the Output format template above to reproduce it.\n\n### Worked example — rollback via PITR (condensed)\n\nCondensed worked example (rollback via PITR with data-loss review items): moved verbatim to [references\u002Fworked-examples.md](references\u002Fworked-examples.md).\nApply the Output format template above to reproduce it.\n\n## Anti-Patterns — NEVER\n\n- NEVER attempt a major version upgrade without a pre-upgrade manual\n  snapshot. The automatic snapshot RDS takes is harder to locate under\n  incident pressure. A named manual snapshot is the rollback anchor.\n\n- NEVER attach a parameter group from the wrong engine family. Aurora\n  MySQL 5.7 param groups are NOT compatible with 8.0 — the instance\n  will boot with defaults, silently dropping custom settings.\n\n- NEVER skip major versions in an upgrade path. PostgreSQL 13 to 15\n  requires going through 14. Aurora MySQL 5.6 to 8.0 requires 5.7\n  first. `describe-db-engine-versions` lists valid targets.\n\n- NEVER upgrade a read replica before its source, or a global cluster\n  secondary before the primary. Always upgrade source\u002Fprimary first.\n\n- NEVER assume `AutoMinorVersionUpgrade: true` handles major versions.\n  It only applies to minor patches within the same major. Major\n  upgrades always require explicit opt-in.\n\n- NEVER run an in-place major upgrade on production without considering\n  blue\u002Fgreen deploy first. In-place upgrades have a 20-minute to 4-hour\n  `upgrading` window. Blue\u002Fgreen provides a sub-60-second switchover\n  with a rollback safety net.\n\n- NEVER delete the blue environment immediately after a blue\u002Fgreen\n  switchover. Keep it 24-72 hours as a rollback safety net.\n\n- NEVER forget to run `ANALYZE` after a PostgreSQL major upgrade.\n  `pg_upgrade` rebuilds planner statistics; without `ANALYZE`, query\n  plans can be 10-100x slower.\n\n- NEVER assume the application driver supports the target version.\n  MySQL 8.0's `caching_sha2_password` breaks older connectors.\n  PostgreSQL 14+ tightened `pg_hba.conf` defaults. Verify BEFORE.\n\n- NEVER rely on \"downgrade\" as rollback. RDS does not support in-place\n  engine downgrades. The ONLY rollback is PITR\u002Fsnapshot restore —\n  which creates a NEW instance with a NEW endpoint.\n\n- NEVER ignore `PendingModifiedValues`. A pending storage, instance\n  class, or parameter change that fires during the upgrade causes\n  unpredictable behavior. Resolve before upgrading.\n\n- NEVER upgrade Aurora Serverless v1 in-place to a major version —\n  restore a snapshot to Serverless v2 instead.\n\n- NEVER auto-execute a state-changing RDS CLI without the CONFIRM gate.\n  Engine upgrades, blue\u002Fgreen switchovers, and PITR restores have\n  downtime and data-loss side effects.\n\n- NEVER overlook `SupportsGlobalDatabases` when upgrading a global\n  cluster member. Not all engine versions support global databases.\n\n## Pre-flight safety checks (run before any remediation CLI)\n\nFour pre-flight safety checks (CONFIRM gate, pre-state capture with snapshot, engine\u002Fparam\u002Foption\u002Fdriver validation, blue\u002Fgreen preference): moved verbatim to [references\u002Fdiagnostic-commands.md](references\u002Fdiagnostic-commands.md).\nDefense-in-depth before any remediation CLI.\n\n## Recent AWS features (2024-2026)\n\nEight recent AWS features 2024-2026 (Blue\u002FGreen GA, Aurora MySQL 3.x auth, PostgreSQL 15, gp3, Serverless v2, global sequencing, RDS Extended Support, Performance Insights retention): moved verbatim to [references\u002Fadvanced-patterns.md](references\u002Fadvanced-patterns.md).\nConsult before citing feature limits or recency.\n\n## References (load on demand)\n\n- [references\u002Fworked-examples.md](references\u002Fworked-examples.md) — Secondary worked examples (blue\u002Fgreen major upgrade, global database upgrade, rollback via PITR).\n- [references\u002Fadvanced-patterns.md](references\u002Fadvanced-patterns.md) — Step 0 non-obvious upgrade behaviors, recent AWS features (2024-2026).\n- [references\u002Ferror-handling.md](references\u002Ferror-handling.md) — Upgrade failure-mode table for diagnosis.\n- [references\u002Fdiagnostic-commands.md](references\u002Fdiagnostic-commands.md) — Live-account pre-flight commands, pre-flight safety checks.\n- [references\u002Fdiagnostic-procedures.md](references\u002Fdiagnostic-procedures.md) — Failure-diagnosis procedures per archetype.\n- [references\u002Fupgrade-paths-and-compatibility.md](references\u002Fupgrade-paths-and-compatibility.md) — Version paths, parameter\u002Foption group families, driver compatibility.\n\n## Domain\n\nAWS CloudOps \u002F RDS and Aurora Engine Upgrade, Major Version Cut-over &\nZero-Downtime Blue\u002FGreen Deploy.\n\n## AWS documentation\n\n- **Amazon RDS User Guide** — https:\u002F\u002Fdocs.aws.amazon.com\u002FAmazonRDS\u002Flatest\u002FUserGuide\u002FWelcome.html\n- **Upgrading a DB instance engine version** — https:\u002F\u002Fdocs.aws.amazon.com\u002FAmazonRDS\u002Flatest\u002FUserGuide\u002FUSER_UpgradeDBInstance.Upgrading.html\n- **Major version upgrades** — https:\u002F\u002Fdocs.aws.amazon.com\u002FAmazonRDS\u002Flatest\u002FUserGuide\u002FUSER_UpgradeDBInstance.MajorVersionUpgrade.html\n- **Minor version upgrades** — https:\u002F\u002Fdocs.aws.amazon.com\u002FAmazonRDS\u002Flatest\u002FUserGuide\u002FUSER_UpgradeDBInstance.MinorVersionUpgrade.html\n- **Blue\u002FGreen Deployments** — https:\u002F\u002Fdocs.aws.amazon.com\u002FAmazonRDS\u002Flatest\u002FUserGuide\u002Fblue-green-deployments.html\n- **Aurora MySQL upgrade paths** — https:\u002F\u002Fdocs.aws.amazon.com\u002FAmazonRDS\u002Flatest\u002FAuroraUserGuide\u002FAuroraMySQL.Updates.Upgrades.html\n- **Aurora PostgreSQL upgrade paths** — https:\u002F\u002Fdocs.aws.amazon.com\u002FAmazonRDS\u002Flatest\u002FAuroraUserGuide\u002FAuroraPostgreSQL.Updates.Upgrades.html\n- **Global database upgrades** — https:\u002F\u002Fdocs.aws.amazon.com\u002FAmazonRDS\u002Flatest\u002FAuroraUserGuide\u002Faurora-global-database-upgrade.html\n- **RDS API Reference** — https:\u002F\u002Fdocs.aws.amazon.com\u002FAmazonRDS\u002Flatest\u002FAPIReference\u002F\n- **Point-in-time recovery** — https:\u002F\u002Fdocs.aws.amazon.com\u002FAmazonRDS\u002Flatest\u002FUserGuide\u002FUSER_PIT.html\n","# RDS and Aurora Engine Upgrade Operator\n\n## What this skill does\n\nExecutes RDS and Aurora engine upgrade operations correctly and safely.\nRuns deterministic pre-checks before any state-changing CLI (instance or\ncluster is `available`, no pending maintenance actions, pre-upgrade\nsnapshot exists and is `available`, parameter group is compatible with the\ntarget engine version, option group is compatible, replication topology\nis understood, storage type and free-space headroom are sufficient),\nexecutes the upgrade behind a CONFIRM gate, and verifies the result by\nconfirming `EngineVersion` matches the target, the instance or cluster\nreturns to `available`, application connectivity is validated, and no\nquery-performance regression exceeds the configured threshold. Every\nmajor upgrade produces an explicit opt-in requirement plus a blue\u002Fgreen\nrecommendation for zero downtime; every minor upgrade surfaces the\n`AutoMinorVersionUpgrade` flag so the operator knows whether the next\npatch is automatic.\n\n## Quick navigation\n\n| Section | What it covers | When to read it |\n|---|---|---|\n| **§ Quick reference** | Verdict thresholds (REVIEW_REQUIRED\u002FOPERATION_COMPLETED) + pre-check priority | Before any operation |\n| **§ Mindset** | Why major upgrades need explicit opt-in, blue\u002Fgreen as the zero-downtime primitive, the param-group trap | Understanding the safety model |\n| **§ Pre-flight** | Database\u002Fcluster metadata gate — state, pending actions, multi-AZ, global cluster, replicas | Before executing any CLI |\n| **§ Process** | Per-operation planning: major, minor, blue-green, param-group, global, auto-minor, verify, rollback | When choosing which operation to run |\n| **§ Output format** | Structured OPERATION\u002FVERDICT\u002FPRE_CHECKS\u002FSTEPS\u002FPOST_VERIFY template | Formatting the response |\n| **§ Anti-Patterns** | NEVER list — common mistakes that cause extended downtime or strand a cluster mid-upgrade | Review before risky operations |\n| **§ Pre-flight safety** | Capture pre-state, snapshot, application driver check, rollback plan | Defense-in-depth |\n\n## Quick reference — verdict thresholds\n\n| Verdict | Trigger condition | Action |\n|---|---|---|\n| `REVIEW_REQUIRED` | Operation plan is ready but requires human review before execution (major upgrade requires explicit opt-in, blue\u002Fgreen plan needs cutover timing approval, global database upgrade sequencing across Regions, PITR rollback plan, parameter group with `pending-reboot` changes that need a maintenance window) | Emit plan with the specific items to review, wait for operator approval |\n| `OPERATION_COMPLETED` | Upgrade finished and post-verification passed (`EngineVersion` matches target, instance or cluster returned to `available`, application connectivity validated, no query regression above threshold, replicas in sync) | Emit verification results, monitoring plan |\n\n**Priority order for pre-checks (apply in this sequence, any failure routes\nto REVIEW_REQUIRED with the failure surfaced):**\n\n1. **Database reachability** — instance or cluster exists, `DBInstanceStatus`\n   or `Status` is `available`, NOT `modifying`, `upgrading`,\n   `backing-up`, or `storage-optimization`. No `PendingModifiedValues`\n   that would conflict with the upgrade.\n2. **Target engine version validity** — `describe-db-engine-versions`\n   confirms the target `EngineVersion` is valid for the instance class,\n   storage type, and current engine. Major version path is supported\n   (skipping major versions is not allowed).\n3. **Pre-upgrade snapshot** — a manual snapshot exists taken AFTER any\n   pending writes, with status `available`. This is the rollback anchor.\n4. **Parameter group compatibility** — a custom parameter group exists for\n   the target engine family (e.g., `default.aurora-mysql8.0` for an Aurora\n   MySQL 5.7 to 8.0 upgrade). The default param group from the source\n   version is NOT compatible.\n5. **Option group compatibility** — an option group for the target engine\n   family exists and includes the same options (or validated replacements).\n6. **Replication topology understood** — read replicas, Aurora replicas,\n   and global cluster secondaries are identified. The upgrade order is\n   primary then replicas (or blue\u002Fgreen switchover).\n7. **Storage headroom** — at least 20% free storage for the upgrade\n   temporary space (engine upgrade can rebuild system tables).\n8. **Application driver support** — the application's database driver\n   version supports the target engine version (e.g., MySQL Connector\u002FJ 8.x\n   for Aurora MySQL 8.0; PostgreSQL JDBC 42.x for PostgreSQL 15).\n9. **Maintenance window** — the upgrade is scheduled in the configured\n   maintenance window OR the operator has explicitly approved an immediate\n   upgrade.\n\n**Cost\u002Ftime baselines (2026):**\n\n- Minor version upgrade: 5-20 minutes of `upgrading` state, brief\n  connection drop on single-AZ; multi-AZ rolls through a failover.\n- Major version upgrade: 20 minutes to 4 hours depending on database\n  size (system table rebuild). Aurora is typically faster than RDS\n  Standard because Aurora stores the system catalogs differently.\n- Blue\u002Fgreen switchover: typically under 60 seconds (DNS shift). The\n  green environment provisioning takes 30-90 minutes before the\n  switchover.\n- Global database upgrade: primary Region first (20 min to 4 hours),\n  then each secondary Region sequentially (the secondary rebuilds from\n  the upgraded primary).\n- PITR rollback to pre-upgrade state: recovery time objective is the\n  snapshot-or-PITR-restore time (minutes to hours depending on size).\n\n## Mindset\n\n**One-line takeaway:** a major version upgrade is an irreversible\ncut-over unless you pre-planned a rollback path (PITR to the pre-upgrade\nmoment or a blue\u002Fgreen switchover back). `modify-db-instance\n--engine-version` returns immediately but the engine upgrade runs\nasynchronously — the database is in `upgrading` state for the full\nduration, during which the old version is gone and the new version is\nnot yet ready. Driven by three RDS realities:\n\n- **Major upgrades require explicit opt-in; auto applies only minor.**\n  `AutoMinorVersionUpgrade: true` schedules minor version patches in the\n  maintenance window. It NEVER triggers a major version upgrade. Setting\n  `--engine-version` to a higher major version is the explicit opt-in\n  and it is irreversible without a restore — you cannot \"downgrade\" an\n  engine in place.\n\n- **Blue\u002Fgreen deploy provides zero-downtime upgrade.** The blue\u002Fgreen\n  pattern creates a full copy of the production environment (the green)\n  running the target engine version, keeps it in sync via logical\n  replication, and switches traffic over via a DNS shift (typically\n  under 60 seconds). This is the recommended path for production major\n  upgrades. The switchover is reversible until the green is deleted.\n\n- **Parameter group must be compatible with the target version.** Each\n  engine family has its own parameter group family (e.g.,\n  `aurora-mysql5.7` vs `aurora-mysql8.0`). An upgrade from 5.7 to 8.0\n  REQUIRES a parameter group from the `aurora-mysql8.0` family. Using\n  the 5.7 parameter group causes the upgrade to fail or the instance to\n  boot with default parameters, silently dropping custom settings like\n  `max_connections`, `innodb_buffer_pool_size`, or `character_set_server`.\n\n## Pre-flight: database metadata gate\n\nRun before classification. Misclassifying these produces wrong plans.\n\n**Pagination:** `describe-db-instances` paginates at 100\u002Fpage — drain\n`--starting-token` to completion.\n\nSix live-account pre-flight commands (instance and cluster metadata, valid upgrade targets, pre-upgrade snapshots, pending maintenance, global topology): moved verbatim to [references\u002Fdiagnostic-commands.md](references\u002Fdiagnostic-commands.md).\nThe attribute table below maps the captured fields to plan effects.\n\n**Malformed input:** if the input JSON is invalid or missing required\nfields, emit `VERDICT: ERROR` with `REASON: DB\u002Fcluster configuration\nis not valid JSON or is missing required fields — cannot plan.` and\n`REMEDIATION: Re-fetch with aws rds describe-db-instances\n--db-instance-identifier \u003Cid> --output json and re-plan.`\n\n| Database attribute | Effect on operation |\n|---|---|\n| `DBInstanceStatus` not `available` | Upgrade BLOCKED. Wait for the current operation to finish, or investigate a stuck status. |\n| `PendingModifiedValues` non-empty | Upgrade routes to REVIEW_REQUIRED. Resolve the pending change (apply immediately or cancel) before scheduling the upgrade. |\n| `MultiAZ: true` | Upgrade runs through a rolling failover. Plan for two brief connection drops (primary -> standby, then upgrade -> failback). |\n| `ReadReplicaSourceDBInstanceIdentifier` set (this is a replica) | Upgrade the SOURCE first, then replicas. Replicas cannot run a higher engine version than the source. |\n| `GlobalClusterIdentifier` set | Global cluster member. Primary Region must upgrade first, then secondaries. Use the global-upgrade operation. |\n| `StorageType: gp2` with \u003C 20% free | Upgrade routes to REVIEW_REQUIRED — increase storage or convert to gp3 first. |\n| `AutoMinorVersionUpgrade: true` | Minor versions are applied automatically in the maintenance window. A manual minor upgrade overrides the schedule. |\n| `DBParameterGroups[0].ParameterApplyStatus: pending-reboot` | A pending parameter change will be applied during the upgrade reboot. Route to REVIEW_REQUIRED. |\n| `DBParameterGroups[0].DBParameterGroupName: default.\u003Cengine>\u003Cversion>` | Default param group. For major upgrades, a custom param group for the target version is required. |\n| `OptionGroupMemberships[0].Status: pending-maintenance` | An option group change is pending. Resolve before upgrade. |\n\n## Process — operation planning (apply in order)\n\n### Step 0: Expert knowledge — non-obvious RDS\u002FAurora upgrade behaviors\n\nTen non-obvious upgrade behaviors (explicit major opt-in, parameter-family migration, blue\u002Fgreen zero-downtime path, multi-AZ rolling failover, apply-immediately semantics, caching_sha2_password, PostgreSQL 14+ statistics, Serverless v1 restore-only, SupportsGlobalDatabases, upgrade-failed recovery): moved verbatim to [references\u002Fadvanced-patterns.md](references\u002Fadvanced-patterns.md).\nLoad before planning any upgrade; Steps 1-4 assume these constraints.\n\n### Step 1: Pre-check gate — REVIEW_REQUIRED if any check needs human attention\n\nRun ALL pre-checks for the chosen operation. If any check fails or\nrequires human review, the verdict is REVIEW_REQUIRED with the specific\nitems listed. Do NOT execute until the operator reviews.\n\n**For ALL operations:**\n1. Instance or cluster exists (`describe-db-instances` \u002F\n   `describe-db-clusters` does not return\n   `DBInstanceNotFound`).\n2. `DBInstanceStatus` \u002F `Status` is `available` (not `modifying`,\n   `upgrading`, `backing-up`, `storage-optimization`, `resetting`,\n   `incompatible-parameters`).\n3. No `PendingModifiedValues` (a pending change conflicts with the\n   upgrade — apply or cancel first).\n4. Storage type and free space: at least 20% free storage. For `gp2`,\n   consider converting to `gp3` first (gp3 supports higher IOPS at\n   lower cost and is the recommended storage type for 2024+).\n5. `PerformanceInsightsEnabled: true` (for before\u002Fafter comparison).\n\n**For major-upgrade (`modify-db-instance --engine-version\n\u003Ctarget-major>`):**\n6. `describe-db-engine-versions` confirms the target version is a\n   valid upgrade target from the current version.\n7. The upgrade path is sequential (no skipped major versions).\n8. A custom parameter group for the target engine family exists and\n   is attached (or will be attached as part of the modify call).\n9. An option group for the target engine family exists and is attached\n   (or will be attached as part of the modify call).\n10. A pre-upgrade manual snapshot exists and is `available`.\n11. Application driver support verified (driver version supports the\n    target engine version).\n12. `PreferredMaintenanceWindow` acknowledged — with\n    `--apply-immediately` the upgrade starts now; without it, the\n    upgrade is deferred to the next window.\n13. Blue\u002Fgreen recommended for production (surface the alternative\n    plan if the operator chose in-place).\n\n**For minor-upgrade (`modify-db-instance --engine-version\n\u003Ctarget-minor>` or `--auto-minor-version-upgrade`):**\n6. `describe-db-engine-versions` confirms the target version is valid.\n7. Minor version is within the same major version.\n8. `AutoMinorVersionUpgrade` current state acknowledged — if `true`\n   and the target minor is the current preferred patch, the upgrade\n   may happen automatically in the next window.\n\n**For blue-green-upgrade (`create-blue-green-deployment`):**\n6. The source instance or cluster is `available`.\n7. The target engine version is a valid upgrade target.\n8. Parameter and option groups for the target version are identified.\n9. The application tolerates logical replication lag during the sync\n   phase (typically seconds; longer for large write-heavy workloads).\n10. A switchover window is planned (the switchover itself is under 60\n    seconds but the green provisioning takes 30-90 minutes).\n\n**For param-group-migrate (pre-stage a target-version parameter group):**\n6. The target parameter group family matches the target engine family\n   (e.g., `aurora-mysql8.0` for Aurora MySQL 8.0).\n7. Parameters diffed against the source param group; custom values\n   reproduced.\n8. Static parameters (those requiring a reboot) identified — they will\n   apply at the next reboot (i.e., during the upgrade).\n\n**For global-upgrade (multi-Region global cluster):**\n6. `describe-global-clusters` confirms the global cluster topology.\n7. The primary Region is identified.\n8. Each secondary Region's cluster is `available`.\n9. The target engine version has `SupportsGlobalDatabases: true`.\n10. A Region-by-Region cutover plan is documented (primary first,\n    secondaries rebuilt sequentially).\n\n**For configure-auto-minor (toggle `AutoMinorVersionUpgrade`):**\n6. The operator understands that enabling auto-minor schedules patches\n   in the maintenance window — it does NOT trigger an immediate patch.\n\n**For post-upgrade-verify (read-only):**\n6. The instance or cluster has returned to `available`.\n7. `EngineVersion` matches the target.\n8. Performance Insights has data for the pre-upgrade and post-upgrade\n   windows.\n\n**For rollback (PITR to pre-upgrade state):**\n6. A pre-upgrade snapshot or a valid PITR window (within the backup\n   retention period) exists.\n7. The operator understands that rollback means a NEW instance or\n   cluster — the endpoint changes. Application connection strings\n   must be updated.\n\nUpgrade failure-mode table (InvalidParameterCombination, parameter\u002Foption group gaps, stuck upgrade-failed, MySQL 8.0 auth plugin, pg_hba.conf, planner-statistics regressions, replica lag, global secondary rebuild): moved verbatim to [references\u002Ferror-handling.md](references\u002Ferror-handling.md).\nLoad when a pre-check fails or an upgrade misbehaves.\n\n### Step 2: OPERATION_COMPLETED or REVIEW_REQUIRED — emit operation plan\n\nIf all pre-checks pass AND the operation is read-only or low-risk\n(minor version with `--apply-immediately`, parameter group migration\nwithout static changes, post-upgrade verification), emit\n`VERDICT: OPERATION_COMPLETED` (for already-finished operations) or\nthe executable plan with the CONFIRM gate.\n\nFor operations requiring human review (major upgrade, blue\u002Fgreen\ncutover, global database sequencing, rollback via PITR), emit\n`VERDICT: REVIEW_REQUIRED` with the specific items to review. The plan\nincludes:\n\n- The exact AWS CLI command with all flags populated from the database\n  configuration.\n- The expected duration (minor 5-20 min; major 20 min to 4 hours;\n  blue\u002Fgreen switchover under 60 seconds after 30-90 min of green\n  provisioning).\n- The expected side-effects (`EngineVersion` changes, instance reboots,\n  connections drop, replicas rebuild).\n- The CONFIRM gate prompt.\n- The rollback plan (snapshot ARN for in-place; blue\u002Fgreen switchback\n  for blue\u002Fgreen).\n\n### Step 3: Execute behind CONFIRM gate\n\n- **MANDATORY CONFIRMATION GATE.** Before any state-changing CLI\n  (`modify-db-instance`, `modify-db-cluster`,\n  `create-blue-green-deployment`, `switchover-blue-green-deployment`,\n  `delete-blue-green-deployment`, `promote-read-replica`,\n  `restore-db-instance-to-point-in-time`), emit:\n  `CONFIRM: About to \u003Coperation> on \u003Cdb-id> in account \u003Caccount>\n  region \u003Cregion>. This will \u003Cconsequence>. Proceed? (yes\u002Fno)`.\n  Do NOT execute until the operator confirms.\n- Capture pre-state for rollback: `aws rds describe-db-instances\n  --db-instance-identifier \u003Cid> --output json > \u002Ftmp\u002F\u003Cid>-pre-$(date\n  +%s).json` AND a manual snapshot if one does not already exist:\n  `aws rds create-db-snapshot --db-instance-identifier \u003Cid>\n  --db-snapshot-id \u003Cid>-pre-upgrade-$(date +%s)`.\n- Execute the CLI. For `modify-db-instance`, the API returns\n  immediately; the actual upgrade runs asynchronously. The instance\n  enters `upgrading` state.\n- Monitor the upgrade via RDS events:\n  `aws rds describe-events --source-type db-instance\n  --source-identifier \u003Cid> --start-time $(($(date +%s) - 600))000`.\n\n### Step 4: Post-verification — OPERATION_COMPLETED\n\nAfter the upgrade finishes, run post-verification. ALL checks must pass\nfor `OPERATION_COMPLETED`.\n\n1. `describe-db-instances --db-instance-identifier \u003Cid>` — confirm\n   `EngineVersion` matches the target AND `DBInstanceStatus: available`.\n2. `describe-db-snapshots` — confirm the automatic snapshot taken by RDS\n   during the upgrade is `available` (this is the post-upgrade backup).\n3. For multi-AZ: confirm both the primary and standby are on the target\n   version (check `SecondaryAvailabilityZone` and failover status).\n4. For Aurora clusters: confirm ALL cluster members (writer and readers)\n   report the target `EngineVersion`. Check `AuroraReplicaLag` has\n   returned to baseline.\n5. For global clusters: confirm ALL secondary Region clusters have been\n   rebuilt and report the target version.\n6. Application connectivity: test a connection from the application's\n   subnet\u002FSG to the database endpoint on the DB port. Verify the\n   application driver version connects successfully.\n7. Query performance verification: compare Performance Insights\n   top-queries latency (post-upgrade) against the pre-upgrade baseline.\n   Investigate any query with > 2x latency regression.\n8. For PostgreSQL: run `ANALYZE` on all user databases to refresh\n   planner statistics. Schedule a `VACUUM ANALYZE` during the next low-\n   traffic window.\n9. For MySQL\u002FAurora MySQL 8.0: verify the authentication plugin\n   (`caching_sha2_password` vs `mysql_native_password`) matches what\n   the application driver expects.\n\nIf ANY verification fails, emit `VERDICT: ERROR` with the failure\ndetails — do not claim OPERATION_COMPLETED. A failed verification\ntypically means a query regression or an application connectivity\nissue; route to the rollback operation if the regression is severe.\n\n## Output format (per operation)\n\n```text\nOPERATION: \u003Cmajor-upgrade | minor-upgrade | blue-green-upgrade | param-group-migrate | global-upgrade | configure-auto-minor | post-upgrade-verify | rollback>\nVERDICT: OPERATION_COMPLETED | REVIEW_REQUIRED\nTARGET: \u003Cdb-id-or-cluster-id> (engine: \u003Ccurrent-version> -> \u003Ctarget-version>)\nPRE_CHECKS:\n  - [PASS] \u003Ccheck description>\n  - [REVIEW] \u003Ccheck description> — \u003Citem to review>\nSTEPS:\n  1. \u003CCLI command with flags populated>\n  2. \u003Cwait \u002F monitoring command>\n  3. \u003Cnext step>\nPOST_VERIFY:\n  - [PASS] \u003Cverification description>\n  - [FAIL] \u003Cverification description> — \u003Creason>\nNOTES: \u003Crollback plan, monitoring, caveats>\n```\n\n### Worked example — minor-upgrade (in-place, OPERATION_COMPLETED)\n\n```text\nOPERATION: minor-upgrade\nVERDICT: OPERATION_COMPLETED\nTARGET: prod-orders-db (engine: postgres 14.10 -> 14.11)\nPRE_CHECKS:\n  - [PASS] DBInstanceStatus: available\n  - [PASS] No PendingModifiedValues\n  - [PASS] describe-db-engine-versions confirms 14.11 is valid for\n    aurora-postgresql on db.r6g.large\n  - [PASS] Storage: 800 GB allocated, 350 GB free (44% headroom)\n  - [PASS] PerformanceInsightsEnabled: true\n  - [PASS] AutoMinorVersionUpgrade: false (operator-initiated)\nSTEPS:\n  1. CONFIRM: About to upgrade prod-orders-db from PostgreSQL 14.10 to\n     14.11 in account 111111111111 region us-east-1. This will reboot\n     the instance (brief connection drop, ~30 seconds for single-AZ,\n     two drops for multi-AZ). Proceed? (yes\u002Fno)\n  2. aws rds create-db-snapshot \\\n       --db-instance-identifier prod-orders-db \\\n       --db-snapshot-id prod-orders-db-pre-14-11-$(date +%s)\n  3. aws rds wait db-snapshot-available \\\n       --db-snapshot-id prod-orders-db-pre-14-11-...\n  4. aws rds modify-db-instance \\\n       --db-instance-identifier prod-orders-db \\\n       --engine-version 14.11 \\\n       --apply-immediately\n  5. aws rds wait db-instance-available \\\n       --db-instance-identifier prod-orders-db\nPOST_VERIFY:\n  - [PASS] EngineVersion: 14.11 (confirmed via describe-db-instances)\n  - [PASS] DBInstanceStatus: available\n  - [PASS] Application connectivity: psql connect from app subnet OK\n  - [PASS] Performance Insights: top-10 query latency within +\u002F- 10%\n    of pre-upgrade baseline\nNOTES:\n  - Minor version upgrade completed in 12 minutes. Single-AZ, one\n    connection drop of ~25 seconds.\n  - PostgreSQL ANALYZE scheduled for the next low-traffic window\n    (03:00 UTC) to refresh planner statistics.\n  - Rollback: restore from snapshot prod-orders-db-pre-14-11-...\n    if any regression is detected. This creates a NEW instance —\n    endpoint will change.\n```\n\n### Worked example — major-upgrade via blue\u002Fgreen (REVIEW_REQUIRED)\n\nFull worked example (Aurora MySQL 5.7 to 8.0 blue\u002Fgreen with driver and switchover review items): moved verbatim to [references\u002Fworked-examples.md](references\u002Fworked-examples.md).\nApply the Output format template above to reproduce it.\n\n### Worked example — global database upgrade (condensed)\n\nCondensed worked example (global database upgrade with Region sequencing): moved verbatim to [references\u002Fworked-examples.md](references\u002Fworked-examples.md).\nApply the Output format template above to reproduce it.\n\n### Worked example — rollback via PITR (condensed)\n\nCondensed worked example (rollback via PITR with data-loss review items): moved verbatim to [references\u002Fworked-examples.md](references\u002Fworked-examples.md).\nApply the Output format template above to reproduce it.\n\n## Anti-Patterns — NEVER\n\n- NEVER attempt a major version upgrade without a pre-upgrade manual\n  snapshot. The automatic snapshot RDS takes is harder to locate under\n  incident pressure. A named manual snapshot is the rollback anchor.\n\n- NEVER attach a parameter group from the wrong engine family. Aurora\n  MySQL 5.7 param groups are NOT compatible with 8.0 — the instance\n  will boot with defaults, silently dropping custom settings.\n\n- NEVER skip major versions in an upgrade path. PostgreSQL 13 to 15\n  requires going through 14. Aurora MySQL 5.6 to 8.0 requires 5.7\n  first. `describe-db-engine-versions` lists valid targets.\n\n- NEVER upgrade a read replica before its source, or a global cluster\n  secondary before the primary. Always upgrade source\u002Fprimary first.\n\n- NEVER assume `AutoMinorVersionUpgrade: true` handles major versions.\n  It only applies to minor patches within the same major. Major\n  upgrades always require explicit opt-in.\n\n- NEVER run an in-place major upgrade on production without considering\n  blue\u002Fgreen deploy first. In-place upgrades have a 20-minute to 4-hour\n  `upgrading` window. Blue\u002Fgreen provides a sub-60-second switchover\n  with a rollback safety net.\n\n- NEVER delete the blue environment immediately after a blue\u002Fgreen\n  switchover. Keep it 24-72 hours as a rollback safety net.\n\n- NEVER forget to run `ANALYZE` after a PostgreSQL major upgrade.\n  `pg_upgrade` rebuilds planner statistics; without `ANALYZE`, query\n  plans can be 10-100x slower.\n\n- NEVER assume the application driver supports the target version.\n  MySQL 8.0's `caching_sha2_password` breaks older connectors.\n  PostgreSQL 14+ tightened `pg_hba.conf` defaults. Verify BEFORE.\n\n- NEVER rely on \"downgrade\" as rollback. RDS does not support in-place\n  engine downgrades. The ONLY rollback is PITR\u002Fsnapshot restore —\n  which creates a NEW instance with a NEW endpoint.\n\n- NEVER ignore `PendingModifiedValues`. A pending storage, instance\n  class, or parameter change that fires during the upgrade causes\n  unpredictable behavior. Resolve before upgrading.\n\n- NEVER upgrade Aurora Serverless v1 in-place to a major version —\n  restore a snapshot to Serverless v2 instead.\n\n- NEVER auto-execute a state-changing RDS CLI without the CONFIRM gate.\n  Engine upgrades, blue\u002Fgreen switchovers, and PITR restores have\n  downtime and data-loss side effects.\n\n- NEVER overlook `SupportsGlobalDatabases` when upgrading a global\n  cluster member. Not all engine versions support global databases.\n\n## Pre-flight safety checks (run before any remediation CLI)\n\nFour pre-flight safety checks (CONFIRM gate, pre-state capture with snapshot, engine\u002Fparam\u002Foption\u002Fdriver validation, blue\u002Fgreen preference): moved verbatim to [references\u002Fdiagnostic-commands.md](references\u002Fdiagnostic-commands.md).\nDefense-in-depth before any remediation CLI.\n\n## Recent AWS features (2024-2026)\n\nEight recent AWS features 2024-2026 (Blue\u002FGreen GA, Aurora MySQL 3.x auth, PostgreSQL 15, gp3, Serverless v2, global sequencing, RDS Extended Support, Performance Insights retention): moved verbatim to [references\u002Fadvanced-patterns.md](references\u002Fadvanced-patterns.md).\nConsult before citing feature limits or recency.\n\n## References (load on demand)\n\n- [references\u002Fworked-examples.md](references\u002Fworked-examples.md) — Secondary worked examples (blue\u002Fgreen major upgrade, global database upgrade, rollback via PITR).\n- [references\u002Fadvanced-patterns.md](references\u002Fadvanced-patterns.md) — Step 0 non-obvious upgrade behaviors, recent AWS features (2024-2026).\n- [references\u002Ferror-handling.md](references\u002Ferror-handling.md) — Upgrade failure-mode table for diagnosis.\n- [references\u002Fdiagnostic-commands.md](references\u002Fdiagnostic-commands.md) — Live-account pre-flight commands, pre-flight safety checks.\n- [references\u002Fdiagnostic-procedures.md](references\u002Fdiagnostic-procedures.md) — Failure-diagnosis procedures per archetype.\n- [references\u002Fupgrade-paths-and-compatibility.md](references\u002Fupgrade-paths-and-compatibility.md) — Version paths, parameter\u002Foption group families, driver compatibility.\n\n## Domain\n\nAWS CloudOps \u002F RDS and Aurora Engine Upgrade, Major Version Cut-over &\nZero-Downtime Blue\u002FGreen Deploy.\n\n## AWS documentation\n\n- **Amazon RDS User Guide** — https:\u002F\u002Fdocs.aws.amazon.com\u002FAmazonRDS\u002Flatest\u002FUserGuide\u002FWelcome.html\n- **Upgrading a DB instance engine version** — https:\u002F\u002Fdocs.aws.amazon.com\u002FAmazonRDS\u002Flatest\u002FUserGuide\u002FUSER_UpgradeDBInstance.Upgrading.html\n- **Major version upgrades** — https:\u002F\u002Fdocs.aws.amazon.com\u002FAmazonRDS\u002Flatest\u002FUserGuide\u002FUSER_UpgradeDBInstance.MajorVersionUpgrade.html\n- **Minor version upgrades** — https:\u002F\u002Fdocs.aws.amazon.com\u002FAmazonRDS\u002Flatest\u002FUserGuide\u002FUSER_UpgradeDBInstance.MinorVersionUpgrade.html\n- **Blue\u002FGreen Deployments** — https:\u002F\u002Fdocs.aws.amazon.com\u002FAmazonRDS\u002Flatest\u002FUserGuide\u002Fblue-green-deployments.html\n- **Aurora MySQL upgrade paths** — https:\u002F\u002Fdocs.aws.amazon.com\u002FAmazonRDS\u002Flatest\u002FAuroraUserGuide\u002FAuroraMySQL.Updates.Upgrades.html\n- **Aurora PostgreSQL upgrade paths** — https:\u002F\u002Fdocs.aws.amazon.com\u002FAmazonRDS\u002Flatest\u002FAuroraUserGuide\u002FAuroraPostgreSQL.Updates.Upgrades.html\n- **Global database upgrades** — https:\u002F\u002Fdocs.aws.amazon.com\u002FAmazonRDS\u002Flatest\u002FAuroraUserGuide\u002Faurora-global-database-upgrade.html\n- **RDS API Reference** — https:\u002F\u002Fdocs.aws.amazon.com\u002FAmazonRDS\u002Flatest\u002FAPIReference\u002F\n- **Point-in-time recovery** — https:\u002F\u002Fdocs.aws.amazon.com\u002FAmazonRDS\u002Flatest\u002FUserGuide\u002FUSER_PIT.html\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, rds, aurora, database, upgrade, mysql, postgresql, blue-green, operate","4","Jacky Chan — AWS Community Builder","aws-cloudops","Databases","0.1.0","RDS upgrade, Aurora upgrade, major version upgrade, minor version upgrade, engine version, auto-minor-version-upgrade, Aurora MySQL 8.0, PostgreSQL 15, blue\u002Fgreen deployment, zero-downtime upgrade, parameter group migration, option group migration, connection string validation, post-upgrade verification, query performance regression, point-in-time recovery, rollback strategy, global database upgrade, multi-AZ upgrade, maintenance window, switchover, UpgradeInProgress","operate","high","false","capability","Upgrading an RDS instance or Aurora cluster engine version (major or minor), planning a major version cut-over with pre-checks and rollback, deploying blue\u002Fgreen for zero-downtime upgrade, migrating parameter or option groups to a target engine version, sequencing a global database upgrade across Regions, scheduling a maintenance-window upgrade, enabling or disabling auto-minor-version-upgrade, validating application connection strings after an upgrade, or verifying post-upgrade query performance and feature parity.","true","OPERATION_COMPLETED | REVIEW_REQUIRED","active","Input: either (a) a database\u002Fcluster configuration (describe-db-instances or describe-db-clusters output) plus the intended operation (major-upgrade, minor-upgrade, blue-green-upgrade, param-group-migrate, global-upgrade, configure-auto-minor, post-upgrade-verify, rollback), OR (b) a DB instance or cluster identifier + operation for live-account execution. Output: deterministic OPERATION \u002F VERDICT \u002F PRE_CHECKS \u002F STEPS \u002F POST_VERIFY \u002F NOTES block per upgrade, where VERDICT is OPERATION_COMPLETED or REVIEW_REQUIRED.","upgrade RDS engine, major version upgrade, minor version upgrade, Aurora MySQL 8.0 upgrade, PostgreSQL 15 upgrade, blue\u002Fgreen deploy RDS, zero-downtime database upgrade, parameter group migration, option group migration, global database upgrade, maintenance window upgrade, auto minor version upgrade, post-upgrade verification, rollback RDS upgrade, RDS engine version","Agent runtime that reads SKILL.md (Claude Code, Cursor, Windsurf, Codex, Gemini). No AWS CLI required for offline plan classification. Live-account operations use aws rds modify-db-instance --engine-version, modify-db-cluster --engine-version, create-blue-green-deployment, switchover-blue-green-deployment, describe-db-engine-versions, describe-db-instances, describe-db-clusters, create-db-snapshot, describe-db-snapshots, modify-db-instance --auto-minor-version-upgrade, promote-read-replica...","main","eeace51bd5dd1db2030729d83eeab988861553aa","skills\u002Frds-upgrade-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 rds-upgrade-operator",{"tool":57,"command":58},"bunx","bunx skills add jhkchan\u002Faws-agent-skills --skill rds-upgrade-operator",{"tool":60,"command":61},"pnpm","pnpm dlx skills add jhkchan\u002Faws-agent-skills --skill rds-upgrade-operator","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Fjhkchan\u002Faws-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Frds-upgrade-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",8335,"https:\u002F\u002Fgithub.com\u002Fjhkchan\u002Faws-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Frds-upgrade-operator\u002Feval\u002Ftest-cases.yaml",{"path":72,"resourceType":67,"resourceTypeDesc":68,"sizeBytes":73,"sourceHtmlUrl":74},"evals\u002Fbaselines\u002Fglobal-database-upgrade-review-required.md",645,"https:\u002F\u002Fgithub.com\u002Fjhkchan\u002Faws-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Frds-upgrade-operator\u002Fevals\u002Fbaselines\u002Fglobal-database-upgrade-review-required.md",{"path":76,"resourceType":67,"resourceTypeDesc":68,"sizeBytes":77,"sourceHtmlUrl":78},"evals\u002Fbaselines\u002Fmajor-upgrade-blue-green-review-required.md",719,"https:\u002F\u002Fgithub.com\u002Fjhkchan\u002Faws-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Frds-upgrade-operator\u002Fevals\u002Fbaselines\u002Fmajor-upgrade-blue-green-review-required.md",{"path":80,"resourceType":67,"resourceTypeDesc":68,"sizeBytes":81,"sourceHtmlUrl":82},"evals\u002Fbaselines\u002Fminor-upgrade-postgres-inplace-completed.md",568,"https:\u002F\u002Fgithub.com\u002Fjhkchan\u002Faws-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Frds-upgrade-operator\u002Fevals\u002Fbaselines\u002Fminor-upgrade-postgres-inplace-completed.md",{"path":84,"resourceType":67,"resourceTypeDesc":68,"sizeBytes":85,"sourceHtmlUrl":86},"evals\u002Fbaselines\u002Fparam-group-missing-review-required.md",556,"https:\u002F\u002Fgithub.com\u002Fjhkchan\u002Faws-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Frds-upgrade-operator\u002Fevals\u002Fbaselines\u002Fparam-group-missing-review-required.md",{"path":88,"resourceType":67,"resourceTypeDesc":68,"sizeBytes":89,"sourceHtmlUrl":90},"evals\u002Fbaselines\u002Frollback-via-pitr-review-required.md",638,"https:\u002F\u002Fgithub.com\u002Fjhkchan\u002Faws-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Frds-upgrade-operator\u002Fevals\u002Fbaselines\u002Frollback-via-pitr-review-required.md",{"path":92,"resourceType":67,"resourceTypeDesc":68,"sizeBytes":93,"sourceHtmlUrl":94},"evals\u002Feval_config.yaml",1959,"https:\u002F\u002Fgithub.com\u002Fjhkchan\u002Faws-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Frds-upgrade-operator\u002Fevals\u002Feval_config.yaml",{"path":96,"resourceType":67,"resourceTypeDesc":68,"sizeBytes":97,"sourceHtmlUrl":98},"evals\u002Fevals.json",4746,"https:\u002F\u002Fgithub.com\u002Fjhkchan\u002Faws-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Frds-upgrade-operator\u002Fevals\u002Fevals.json",{"path":100,"resourceType":67,"resourceTypeDesc":68,"sizeBytes":101,"sourceHtmlUrl":102},"evals\u002Fprompts\u002Fglobal-database-upgrade-review-required.md",980,"https:\u002F\u002Fgithub.com\u002Fjhkchan\u002Faws-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Frds-upgrade-operator\u002Fevals\u002Fprompts\u002Fglobal-database-upgrade-review-required.md",{"path":104,"resourceType":67,"resourceTypeDesc":68,"sizeBytes":105,"sourceHtmlUrl":106},"evals\u002Fprompts\u002Fmajor-upgrade-blue-green-review-required.md",1016,"https:\u002F\u002Fgithub.com\u002Fjhkchan\u002Faws-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Frds-upgrade-operator\u002Fevals\u002Fprompts\u002Fmajor-upgrade-blue-green-review-required.md",{"path":108,"resourceType":67,"resourceTypeDesc":68,"sizeBytes":109,"sourceHtmlUrl":110},"evals\u002Fprompts\u002Fminor-upgrade-postgres-inplace-completed.md",1060,"https:\u002F\u002Fgithub.com\u002Fjhkchan\u002Faws-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Frds-upgrade-operator\u002Fevals\u002Fprompts\u002Fminor-upgrade-postgres-inplace-completed.md",{"path":112,"resourceType":67,"resourceTypeDesc":68,"sizeBytes":113,"sourceHtmlUrl":114},"evals\u002Fprompts\u002Fparam-group-missing-review-required.md",999,"https:\u002F\u002Fgithub.com\u002Fjhkchan\u002Faws-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Frds-upgrade-operator\u002Fevals\u002Fprompts\u002Fparam-group-missing-review-required.md",{"path":116,"resourceType":67,"resourceTypeDesc":68,"sizeBytes":117,"sourceHtmlUrl":118},"evals\u002Fprompts\u002Frollback-via-pitr-review-required.md",927,"https:\u002F\u002Fgithub.com\u002Fjhkchan\u002Faws-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Frds-upgrade-operator\u002Fevals\u002Fprompts\u002Frollback-via-pitr-review-required.md",{"path":120,"resourceType":67,"resourceTypeDesc":68,"sizeBytes":121,"sourceHtmlUrl":122},"examples\u002FREADME.md",9100,"https:\u002F\u002Fgithub.com\u002Fjhkchan\u002Faws-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Frds-upgrade-operator\u002Fexamples\u002FREADME.md",{"path":124,"resourceType":67,"resourceTypeDesc":68,"sizeBytes":125,"sourceHtmlUrl":126},"references\u002Fadvanced-patterns.md",4515,"https:\u002F\u002Fgithub.com\u002Fjhkchan\u002Faws-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Frds-upgrade-operator\u002Freferences\u002Fadvanced-patterns.md",{"path":128,"resourceType":67,"resourceTypeDesc":68,"sizeBytes":129,"sourceHtmlUrl":130},"references\u002Fdiagnostic-commands.md",2596,"https:\u002F\u002Fgithub.com\u002Fjhkchan\u002Faws-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Frds-upgrade-operator\u002Freferences\u002Fdiagnostic-commands.md",{"path":132,"resourceType":67,"resourceTypeDesc":68,"sizeBytes":133,"sourceHtmlUrl":134},"references\u002Fdiagnostic-procedures.md",11705,"https:\u002F\u002Fgithub.com\u002Fjhkchan\u002Faws-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Frds-upgrade-operator\u002Freferences\u002Fdiagnostic-procedures.md",{"path":136,"resourceType":67,"resourceTypeDesc":68,"sizeBytes":137,"sourceHtmlUrl":138},"references\u002Ferror-handling.md",1992,"https:\u002F\u002Fgithub.com\u002Fjhkchan\u002Faws-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Frds-upgrade-operator\u002Freferences\u002Ferror-handling.md",{"path":140,"resourceType":67,"resourceTypeDesc":68,"sizeBytes":141,"sourceHtmlUrl":142},"references\u002Fupgrade-paths-and-compatibility.md",6507,"https:\u002F\u002Fgithub.com\u002Fjhkchan\u002Faws-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Frds-upgrade-operator\u002Freferences\u002Fupgrade-paths-and-compatibility.md",{"path":144,"resourceType":67,"resourceTypeDesc":68,"sizeBytes":145,"sourceHtmlUrl":146},"references\u002Fworked-examples.md",5825,"https:\u002F\u002Fgithub.com\u002Fjhkchan\u002Faws-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Frds-upgrade-operator\u002Freferences\u002Fworked-examples.md",{"path":148,"resourceType":67,"resourceTypeDesc":68,"sizeBytes":149,"sourceHtmlUrl":150},"skill.usf.yaml",4701,"https:\u002F\u002Fgithub.com\u002Fjhkchan\u002Faws-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Frds-upgrade-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",[],[]]