AGENT SKILL

migration-hub-strategy-deployer

Deploys AWS Migration Hub Strategy Recommendations assessments: assessment creation and group-by-application analysis, data collection via Application Discovery Service (agent-based for dependency telemetry vs agentless Collector VM for inventory), the 6R strategy classification (rehost, replatform, refactor, repurchase, retain, retire) from runtime telemetry, anti-pattern detection (EOL OS, deprecated DB engines), TCO analysis, right-sizing from utilization percentiles, migration wave planning from dependency mapping, and handoff to Application Migration Service (MGN) for rehost and Database Migration Service (DMS) for replatform. Emits READY_TO_DEPLOY with a Collector deployment plan and assessment schedule or PREREQUISITES_MISSING with the specific gap. Use when planning a migration assessment, deploying the Collector, interpreting 6R results, or sequencing migration waves.

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

01 前言 Front Matter

name
migration-hub-strategy-deployer
license
Apache-2.0
metadata
{ "tags": "aws-migration-hub, strategy-recommendations, application-discovery-service, collector, 6r-strategy, wave-planning, mgn, dms, deploy", "phase": "1", "author": "Jacky Chan — AWS Community Builder", "domain": "aws-cloudops", "family": "Migration", "version": "0.1.0", "keywords": "AWS Migration Hub, Strategy Recommendations, Application Discovery Service, ADS, Agent-based discovery, Agentless discovery, Collector VM, vCenter, 6R strategy, rehost, replatform, refactor, repurchase, retain, retire, anti-pattern detection, end-of-life OS, deprecated DB engine, TCO analysis, right-sizing, migration wave planning, dependency mapping, Application Migration Service, MGN, Database Migration Service, DMS", "task_type": "deploy", "complexity": "high", "entry_point": "false", "skill_class": "capability", "when_to_use": "Planning a Migration Hub Strategy Recommendations assessment, deploying the Application Discovery Service Collector (agent-based or agentless), interpreting 6R strategy classifications from runtime telemetry, detecting anti-patterns (end-of-life OS, deprecated DB engines), running a TCO analysis, right-sizing instances from utilization data, planning migration waves from dependency maps, or handing off to Application Migration Service (MGN) and Database Migration Service (DMS) for execution.", "requires_llm": "true", "verdict_shape": "READY_TO_DEPLOY | PREREQUISITES_MISSING", "lifecycle_status": "active", "invocation_schema": "Input shape (one of): (a) a migration assessment specification including data source (agent-based, agentless import, or existing Discovery data), target applications or servers, assessment duration, and strategy preferences; (b) a partial spec for interactive refinement (\"assess 200 vCenter VMs for 6R strategy\"); (c) existing assessment results for interpretation and wave planning. Output shape: { ASSESSMENT_SPEC, VERDICT, COLLECTOR_PLAN, ASSESSMENT_SCHEDULE, STRATEGY_GROUPS[], ANTI_PATTERNS[], WAVE_PLAN, FINDINGS[] } where VERDICT is READY_TO_DEPLOY (Collector plan and assessment ready) or PREREQUISITES_MISSING (specific gap cited).", "supports_pipeline": "true", "activation_triggers": "create Migration Hub Strategy assessment, deploy Discovery Collector, agent-based vs agentless discovery, 6R strategy classification, rehost replatform refactor repurchase retain retire, anti-pattern detection end-of-life OS, deprecated DB engine migration, TCO analysis migration, right-sizing recommendations, migration wave planning, dependency mapping migration, Application Migration Service MGN handoff, Database Migration Service DMS replatform" }
description
Deploys AWS Migration Hub Strategy Recommendations assessments: assessment creation and group-by-application analysis, data collection via Application Discovery Service (agent-based for dependency telemetry vs agentless Collector VM for inventory), the 6R strategy classification (rehost, replatform, refactor, repurchase, retain, retire) from runtime telemetry, anti-pattern detection (EOL OS, deprecated DB engines), TCO analysis, right-sizing from utilization percentiles, migration wave planning from dependency mapping, and handoff to Application Migration Service (MGN) for rehost and Database Migration Service (DMS) for replatform. Emits READY_TO_DEPLOY with a Collector deployment plan and assessment schedule or PREREQUISITES_MISSING with the specific gap. Use when planning a migration assessment, deploying the Collector, interpreting 6R results, or sequencing migration waves.
compatibility
Agent runtime that reads SKILL.md (Claude Code, Cursor, Windsurf, Codex, Gemini). No AWS CLI required for offline assessment planning. Live deployment uses aws migrationhub-strategy start-assessment, get-assessment, list-assessment, aws discovery describe-agents, describe-configurations, list-configurations, start-import-task, get-import-task, aws migrationhuborchestrator create-workflow, and aws mgn create-launch-template-template — AWS CLI v2, SSO or key-based credentials, Discovery and...

02 资源文件 19

03 SKILL.md 正文

Migration Hub Strategy Deployer

Mindset

One-line takeaway: a migration strategy assessment is a three-stage pipeline — collect (Application Discovery Service gathers inventory, performance, and network dependency telemetry via agent-based Collectors or agentless import) → classify (Strategy Recommendations derives the 6R strategy per application from runtime data and anti-pattern detection) → sequence (wave planning groups applications by dependency proximity and risk into deployable migration waves). A gap in ANY stage produces a silent failure: the Collector is deployed but never reports data, the assessment runs but classifies everything as "rehost" because utilization data is missing, or the wave plan ignores a hard dependency and the first wave breaks a downstream application.

→ Extended Mindset rationale moved verbatim to references/advanced-patterns.md.

Quick navigation

You want to... Go to
Pick between agent-based and agentless discovery Step 1 + trade-off table
Deploy the Collector VM on vCenter Step 2 + worked CLI
Configure the assessment scope and duration Step 3
Interpret the 6R strategy classification Step 4 + Appendix A
Detect anti-patterns (EOL OS, deprecated DB) Step 5
Run TCO analysis and right-sizing Step 6
Plan migration waves from dependency data Step 7
Hand off to MGN for rehost execution Step 8
Hand off to DMS for database replatform Step 9
Avoid common Collector and assessment pitfalls Anti-Patterns
Recent features (Strategy Recommendations API, MGN wave builder) Recent AWS features

Critical rules at a glance (do NOT bury these)

  1. Agent-based discovery is required for dependency mapping. Agentless (Collector VM) gathers inventory and basic performance but CANNOT capture inter-host network connections. If the wave plan depends on dependency data, agents must be installed on all in-scope hosts before the assessment starts.
  2. Strategy Recommendations needs 7+ days of utilization data for right-sizing. The assessment derives EC2 instance recommendations from utilization percentiles. A 1-day collection produces noisy recommendations; the minimum recommended window is 7 days, with 14-30 days preferred for seasonal workloads.
  3. The Discovery Collector VM requires network connectivity to BOTH the vCenter and AWS (the home region). A Collector that cannot reach vCenter cannot import inventory; a Collector that cannot reach AWS cannot upload data. Both firewalls must be open before the Collector starts.
  4. MGN replication requires the source server's root volume to be accessible and the MGN agent installed. A rehost strategy is not "done" when the 6R says rehost — it is done when MGN replication is healthy and a test cutover succeeds.
  5. DMS replatform requires the source and target database engines to be compatible. Strategy Recommendations may recommend replatform from Oracle to Aurora PostgreSQL, but DMS schema conversion (via AWS SCT) is a prerequisite. The assessment does not run SCT automatically.

Pre-flight: data requirements

Planning a Migration Hub Strategy assessment requires these inputs:

Input Source Why
Home region aws migrationhub-strategy get-portfolio-preferences (region-aware) Migration Hub is region-scoped; all data lands in the home region
Existing Discovery data aws discovery describe-agents / list-configurations Avoid re-collecting if agents are already deployed
vCenter connectivity Network firewall rules Collector VM must reach vCenter on 443
Application inventory CMDB, spreadsheet, or vCenter VM list Defines the assessment scope
MGN service role iam get-role on AWSApplicationMigrationAgentInstallationRole Rehost handoff requires MGN to be set up
DMS endpoints aws dms describe-endpoints Database replatform requires source and target endpoints
Strategy preferences migrationhub-strategy get-portfolio-preferences Configure rehost/replatform/refactor weighting

If the input is malformed (missing vCenter or agent list, ambiguous home region), emit:

ASSESSMENT: <reference>
VERDICT: ERROR
REASON: Cannot plan assessment — data source and application inventory are required.
GAP: Re-supply the discovery method (agent-based / agentless), the application or server list, and the target home region.

Process — Assessment planning (apply in order, produce deployment plan)

Step 0: Expert knowledge — non-obvious Discovery and Strategy behaviors

→ Step 0 expert-knowledge deep dive moved verbatim to references/advanced-patterns.md.

Step 1: Choose the discovery method

Dimension Agent-based (Discovery Agent) Agentless (Collector VM)
Installation Agent on each guest OS OVA on vCenter or hypervisor
Inventory data Yes Yes
Performance data Process-level (CPU, memory, disk per process) VM-level (CPU, memory, disk for the whole VM)
Network dependency mapping Yes (process-to-process connections) No (VM-level only)
Guest OS access required Yes (install agent) No
Reboot required Sometimes (driver install on Windows) No
Best for Dependency mapping, right-sizing, application grouping Quick inventory, vCenter-only environments
Minimum collection window for 6R 7 days 7 days (but 6R falls back to rehost without dependencies)

Decision rule: deploy agent-based discovery on all critical applications (those with inter-host dependencies, databases, or right-sizing uncertainty). Deploy agentless for the long tail of standalone VMs where inventory-only is sufficient. The two data sources merge in Discovery.

Step 2: Deploy the Discovery Collector (agentless)

Download the Collector OVA from the Migration Hub console and deploy on vCenter:

→ Collector registration CLI moved verbatim to references/discovery-collector-deployment.md.

Collector prerequisites checklist:

Prerequisite How to verify Effect if missing
vCenter reachability (443) nc -zv vcenter.example.com 443 Collector cannot import inventory
AWS reachability (443 to home region) nc -zv migrationhub-strategy.us-east-1.amazonaws.com 443 Collector cannot upload data
vCenter credentials with read access Test login Collector cannot enumerate VMs
Collector VM network with DHCP or static IP vCenter console Collector cannot communicate
DNS resolution for AWS endpoints nslookup migrationhub-strategy.us-east-1.amazonaws.com Upload fails intermittently

Step 3: Configure and start the assessment

# Start the assessment (assumes Discovery data is already collected)
aws migrationhub-strategy start-assessment \
  --assessment-name portfolio-assessment-q3 \
  --s3bucket-for-report-data migration-reports-111111111111 \
  --region us-east-1

Assessment scope and duration:

Parameter Recommended value Why
Collection window before assessment 7-14 days Right-sizing needs percentile data
Assessment scope All in-scope servers Group-by-application requires full scope
S3 report bucket Dedicated bucket with lifecycle policy Reports are large; set 90-day expiration
Strategy preferences Configure before starting rehost vs replatform weighting affects output

Step 4: Interpret the 6R strategy classification

Strategy Recommendations assigns one of six strategies per application:

Strategy Meaning When it is recommended
Rehost Lift-and-shift to EC2 with minimal changes Default when no anti-patterns detected and no replatform driver
Replatform Move to a managed AWS service with minor changes (e.g., EC2 to Elastic Beanstalk, self-managed DB to RDS) When a managed service reduces operational overhead without code changes
Refactor Redesign the application for cloud-native architecture When the application has a strong cloud-native driver (scalability, cost)
Repurchase Move to a SaaS or different product When the current product is being replaced (e.g., commercial CRM to SaaS)
Retain Keep on-premises or defer migration When the application is sunsetting soon or has hard compliance constraints
Retire Decommission the application When the application is unused (Discovery shows zero activity)

Key insight: the 6R is a starting point, not a mandate. Operators should validate each recommendation against business context. A rehost recommendation for a stateful monolith may be technically correct but strategically wrong if the business wants to refactor for scalability.

Step 5: Detect anti-patterns

Strategy Recommendations flags anti-patterns that affect the strategy:

Anti-pattern Detection signal Strategy impact
End-of-life OS (e.g., Windows Server 2012 R2, CentOS 7) OS version from agent or vCenter Forces replatform (upgrade OS) or refactor (containerize)
Deprecated DB engine (e.g., Oracle 11g, SQL Server 2008) DB engine version from agent Forces replatform (upgrade DB) or repurchase (move to managed)
Kernel version gap Linux kernel version May block MGN agent compatibility; update required
Bare-metal host vCenter host type Rehost not possible; must virtualize first
Excessive resource allocation Utilization < 10% peak Right-sizing recommendation; rehost with smaller instance
High network dependency (100+ connections) Agent network data Wave planning must sequence this application carefully

Step 6: TCO analysis and right-sizing

Strategy Recommendations generates a TCO comparison between on-premises and AWS, based on:

  • On-premises hardware cost (from the operator's input or defaults)
  • AWS cost (from right-sized EC2 recommendations + EBS + data transfer)
  • 3-year TCO horizon (default)

Right-sizing methodology:

Utilization percentile EC2 recommendation
P95 CPU < 20%, memory < 30% Downsize by 2 sizes
P95 CPU 20-60%, memory 30-70% Current size or one size down
P95 CPU > 80% or memory > 90% Upsize or split

Gap class — missing utilization data: if the collection window is less than 7 days or the agent was not installed, right-sizing falls back to a match-current-spec recommendation (on-premises vCPU/RAM mapped to the closest EC2 instance). This over-provisions in most cases. Flag PREREQUISITES_MISSING if the collection window is insufficient.

Step 7: Plan migration waves from dependency data

Wave planning groups applications by dependency proximity. The goal is to migrate interdependent applications together to avoid partial-migration breakage.

Wave planning inputs:

Input Source
Application dependency graph Discovery agent network data
Application grouping Strategy Recommendations group-by-application output
Business priority Stakeholder input
Risk tolerance Stakeholder input
Cutover window constraints Stakeholder input

Wave sequencing heuristic:

  1. Wave 0 (pilot): 1-2 low-risk, standalone applications (no dependencies). Validate the MGN/DMS pipeline end-to-end.
  2. Wave 1 (foundation): shared services (DNS, AD, monitoring) that downstream applications depend on. These MUST migrate first.
  3. Wave 2+ (dependency-ordered): applications grouped by dependency clusters. Each wave is a connected component in the dependency graph — migrating a wave must not leave a dependency stranded across the on-premises/AWS boundary.
  4. Wave N (complex): applications with 100+ dependencies or hard compliance constraints. These require detailed runbooks and may use a hybrid model (retain on-premises with VPN/Direct Connect).

Hard constraint: a wave must not split a dependency cluster. If application A (wave 2) depends on application B (wave 3), either move B to wave 2 or delay A to wave 3. The dependency graph determines the minimum wave count; business priority determines the ordering within that constraint.

Step 8: Hand off to Application Migration Service (MGN) for rehost

For applications classified as "rehost," MGN handles replication and cutover:

# Install the MGN agent on the source server (requires OS access)
# On Linux:
sudo yum install aws-replication-agent -y

# Register the source server with MGN
aws mgn create-launch-template-template \
  --launch-template-template-name rehost-template \
  --region us-east-1

MGN handoff checklist:

Step Command / action
Install MGN agent Per OS (yum, apt, MSI)
Verify replication healthy aws mgn describe-replication-configuration-templates
Configure launch template Set instance type, subnet, security group
Run test cutover aws mgn start-cutover --source-server-id s-xxxx
Validate test instance Application-level smoke tests
Schedule production cutover Stakeholder approval + maintenance window

Step 9: Hand off to Database Migration Service (DMS) for replatform

For databases classified as "replatform" (self-managed to RDS/Aurora):

# Create the DMS replication instance
aws dms create-replication-instance \
  --replication-instance-identifier migration-rep \
  --replication-instance-class dms.r5.xlarge \
  --region us-east-1

DMS handoff checklist:

Step Command / action
Run AWS SCT for schema conversion SCT GUI or CLI (heterogeneous migrations only)
Create source endpoint aws dms create-endpoint --endpoint-type source ...
Create target endpoint aws dms create-endpoint --endpoint-type target ...
Create migration task aws dms create-replication-task ...
Start full load + CDC aws dms start-replication-task --migration-type full-load-and-cdc
Validate data parity Compare row counts and checksums
Cutover application connection Update connection strings

Output format

ASSESSMENT: <reference>
HOME_REGION: <region>
DISCOVERY_METHOD:
  - Agent-based: <yes | no>, agent count <N>
  - Agentless: <yes | no>, Collector VM on <vCenter / hypervisor>
  - Collection window: <days>
COLLECTOR_PLAN:
  - Prerequisites: <verified list>
  - Deployment: <OVA + registration steps>
ASSESSMENT_SCHEDULE:
  - Start: <immediate | after-collection>
  - Duration: <days>
  - Report bucket: <s3 uri>
STRATEGY_GROUPS:
  - Application: <name>, servers: <N>, strategy: <6R>, confidence: <high | medium | low>
ANTI_PATTERNS:
  - Server: <id>, pattern: <EOL OS | deprecated DB | ...>, impact: <replatform | refactor | ...>
TCO:
  - On-premises 3-year: <$X>
  - AWS 3-year (right-sized): <$Y>
  - Savings: <$Z>
RIGHT_SIZING:
  - Server: <id>, current: <vCPU/RAM>, recommended: <instance type>, utilization: <P95 CPU/mem>
WAVE_PLAN:
  - Wave 0: <pilot applications>
  - Wave 1: <foundation applications>
  - Wave 2+: <dependency-ordered clusters>
MGN_HANDOFF: <ready | blocked>
DMS_HANDOFF: <ready | blocked>
VERDICT: READY_TO_DEPLOY | PREREQUISITES_MISSING
GAP: <if PREREQUISITES_MISSING, the specific missing piece>
TEMPLATE: <CLI snippets for Collector, assessment, MGN, DMS>

Worked example — READY_TO_DEPLOY, full assessment

ASSESSMENT: portfolio-q3-assessment
HOME_REGION: us-east-1
DISCOVERY_METHOD:
  - Agent-based: yes, agent count 120 (all critical hosts)
  - Agentless: yes, Collector VM on vcenter.example.com
  - Collection window: 14 days
COLLECTOR_PLAN:
  - Prerequisites: vCenter 443 reachable, AWS 443 reachable, vCenter read credentials verified, DNS resolving
  - Deployment: OVA deployed, registered to home region us-east-1
ASSESSMENT_SCHEDULE:
  - Start: after-collection (14 days from Collector deploy)
  - Duration: 14 days collection + 1 day analysis
  - Report bucket: s3://migration-reports-111111111111/portfolio-q3/
STRATEGY_GROUPS:
  - Application: billing-service, servers: 8, strategy: rehost, confidence: high
  - Application: user-auth, servers: 5, strategy: replatform (self-managed PostgreSQL to RDS), confidence: high
  - Application: legacy-crm, servers: 12, strategy: refactor (containerize), confidence: medium
  - Application: reporting-etl, servers: 0 active, strategy: retire, confidence: high
ANTI_PATTERNS:
  - Server: srv-042, pattern: EOL OS (Windows Server 2012 R2), impact: replatform to Windows Server 2022 required
  - Server: srv-078, pattern: deprecated DB engine (Oracle 11g), impact: replatform to Aurora PostgreSQL or upgrade
  - Server: srv-091, pattern: excessive resource allocation (P95 CPU 3%), impact: right-size from m5.4xlarge to m5.large
TCO:
  - On-premises 3-year: $1,800,000
  - AWS 3-year (right-sized): $1,100,000
  - Savings: $700,000 (39%)
RIGHT_SIZING:
  - Server: srv-091, current: 16 vCPU / 64 GB, recommended: m5.large (2 vCPU / 8 GB), utilization: P95 CPU 3%, mem 12%
  - Server: srv-042, current: 8 vCPU / 32 GB, recommended: m5.xlarge (4 vCPU / 16 GB), utilization: P95 CPU 45%, mem 60%
WAVE_PLAN:
  - Wave 0 (pilot): billing-service (8 servers, no external dependencies)
  - Wave 1 (foundation): shared-dns, shared-ad, shared-monitoring (15 servers, all downstream depend on these)
  - Wave 2: user-auth (5 servers, depends on shared-ad from Wave 1)
  - Wave 3: legacy-crm (12 servers, depends on shared-dns and user-auth)
MGN_HANDOFF: ready (agent install instructions prepared for rehost targets)
DMS_HANDOFF: ready (SCT + DMS endpoints prepared for user-auth PostgreSQL replatform)
VERDICT: READY_TO_DEPLOY
GAP: None
TEMPLATE:
  # 1. Deploy Collector VM (download OVA from console, deploy on vcenter.example.com)
  aws discovery start-import-task --name vcenter-import --vcenter --region us-east-1

  # 2. Install agents on 120 critical hosts (yum/apt/MSI per OS)

  # 3. Wait 14 days, then start assessment
  aws migrationhub-strategy start-assessment --assessment-name portfolio-q3-assessment --s3bucket-for-report-data migration-reports-111111111111 --region us-east-1

  # 4. MGN rehost for billing-service
  aws mgn create-launch-template-template --launch-template-template-name billing-rehost --region us-east-1
  # 5. DMS replatform for user-auth PostgreSQL
  aws dms create-replication-instance --replication-instance-identifier user-auth-rep --replication-instance-class dms.r5.xlarge --region us-east-1

Worked example — PREREQUISITES_MISSING, agentless-only with short window

→ Secondary worked example moved verbatim to references/worked-examples.md.

Anti-Patterns — NEVER do these things

  • NEVER start an assessment with less than 7 days of Discovery data. Right-sizing and 6R classification rely on utilization percentiles. A short window produces noisy recommendations.

  • NEVER rely on agentless-only discovery for wave planning. The Collector VM provides inventory but NOT inter-host dependencies. Without the dependency graph, wave sequencing is impossible. Install agents on critical hosts.

  • NEVER rehost an end-of-life OS without flagging the anti-pattern. Strategy Recommendations detects EOL OS and recommends replatform. Ignoring this produces an EC2 instance that fails compliance on day one.

  • NEVER assume the 6R strategy is final. It is derived from runtime telemetry, not business context. Validate each recommendation with stakeholders.

  • NEVER skip the MGN test cutover. A rehost is not complete until MGN replication is healthy AND a test cutover produces a bootable, functional EC2 instance.

  • NEVER skip AWS SCT for heterogeneous database migrations. DMS replicates data; it does NOT convert schema objects or stored procedures. An Oracle-to-Aurora replatform without SCT produces a database with tables but no triggers, views, or procedures.

  • NEVER deploy the Collector VM without verifying vCenter AND AWS network reachability (both paths on 443). A Collector that cannot reach either endpoint appears healthy but never reports data.

  • NEVER change the Migration Hub home region after collection starts. All Discovery data is region-scoped. Confirm the region with stakeholders before starting.

  • NEVER wave-plan without the dependency graph. A wave that splits a dependency cluster creates a broken state where applications cannot reach each other across the on-premises/AWS boundary.

  • NEVER start DMS replication without validating data parity after the full load. Compare row counts and checksums before cutover.

  • NEVER interpret "retain" as "do nothing." Retained applications still require a connectivity plan (VPN/Direct Connect) to reach migrated applications in AWS.

  • NEVER use Migration Hub Orchestrator managed templates without verifying they fit the application profile. Validate against one pilot application first.

Configuration dependency graph

→ Configuration dependency graph moved verbatim to references/advanced-patterns.md.

Pre-flight safety checks

  • MANDATORY CONFIRMATION GATE. Before any state-changing operation, emit: CONFIRM: About to <action> for assessment <name> in <region>. Proceed? (yes/no)
  • Before deploying the Collector VM, verify vCenter and AWS network paths are open.
  • Before starting the assessment, verify the collection window is at least 7 days.
  • Before a production MGN cutover, verify a test cutover succeeded.
  • Before a DMS cutover, verify data parity (row counts + checksums).

Appendix A — 6R strategy decision matrix

→ Appendix A decision matrix moved verbatim to references/strategy-and-wave-planning.md.

Appendix B — Collector VM vs Discovery Agent deployment summary

→ Appendix B comparison table moved verbatim to references/discovery-collector-deployment.md.

For the full Collector deployment walkthrough and agent install commands, see references/discovery-collector-deployment.md. For the 6R interpretation guide and wave planning templates, see references/strategy-and-wave-planning.md.

Recent AWS features (2024-2026)

→ Recent AWS features deep dive moved verbatim to references/advanced-patterns.md.

Expert heuristic: agent-based vs agentless trade-off + 6R from telemetry + wave sequencing

→ Expert-heuristic deep dive moved verbatim to references/advanced-patterns.md.

References (load on demand)

Domain

AWS CloudOps / Migration — Strategy assessment and wave planning.

AWS documentation

04 同仓库更多 Skills

jhkchan
jhkchan/aws-agent-skills

xray-tracing-deployer

Deploys AWS X-Ray distributed tracing with production-grade configuration: X-Ray daemon deployment (EC2 systemd, ECS sidecar, EKS DaemonSet, Lambda built-in), SDK instrumentation (Java, Python, Node.js, Go, .NET) with annotations and metadata, sampling rules (default reservoir + rate vs custom rules with URL / method / service-name predicates), service map generation, trace query and retrieval, groups and insights, CloudWatch ServiceLens integration, IAM permissions (PutTraceSegments, GetSamplingRules), and latest features (Lambda Powertools tracing, OpenTelemetry / ADOT Collector support, W3C trace context). Emits a READY_TO_DEPLOY checklist with every configuration item verified. Use when enabling X-Ray on a new application, adding tracing to ECS / EKS / Lambda, validating a sampling configuration, or generating daemon deployment CLI and IaC templates. Triggers: X-Ray, distributed tracing, X-Ray daemon, sampling rules, service map, X-Ray SDK, OpenTelemetry, ADOT, Lambda Powertools, CloudWatch ServiceLens.

★ 4
jhkchan
jhkchan/aws-agent-skills

wellarchitected-workload-auditor

Audits AWS Well-Architected Tool workloads for review staleness, per-pillar high-risk issue counts, milestone tracking gaps, and remediation plan completeness. Emits a deterministic verdict (STALE_REVIEW | HIGH_RISK | CONFIG_GAP | OK) per workload with per-pillar risk breakdown, milestone trend analysis, and specific remediation. Use when reviewing Well-Architected workloads, checking WA review freshness, auditing high-risk issues per pillar, validating milestone coverage, or tracking remediation plans.

★ 4
jhkchan
jhkchan/aws-agent-skills

wellarchitected-review-operator

Operates AWS Well-Architected Tool reviews end-to-end — creates workloads, runs pillar reviews (operational excellence, security, reliability, performance, cost optimization, sustainability, Prosperity), answers questions with risk-tier rationale, generates reports and improvement plans, creates milestones, and integrates with Trusted Advisor and Well-Architected Labs. Runs deterministic pre-checks (workload existence, lens availability, pillar coverage), emits the exact update-answer / get-consolidated-report CLI behind a CONFIRM gate, and verifies risk-tier transitions. Emits READY | BLOCKED | COMPLETED. Use when creating a workload, answering pillar questions, generating an improvement plan, creating a milestone, or integrating with Trusted Advisor. Triggers: well-architected tool, well architected review, workload review, pillar, operational excellence, security, reliability, performance, cost optimization, sustainability, prosperity, improvement plan, milestone, trusted advisor.

★ 4
jhkchan
jhkchan/aws-agent-skills

wafv2-web-acl-deployer

Provisions AWS WAFv2 Web ACLs with correct production defaults: scope selection (CLOUDFRONT vs REGIONAL), managed rule groups (AWSManagedRulesCommonRuleSet, SQLiRuleSet, LinuxRuleSet, WindowsRuleSet, IPReputationList, BotControlRuleSet, AmazonIpReputationList, ATP), custom rules (byte match, IP set, regex pattern, geo match, size constraint, rate-based), rule priority ordering, logging destinations (Kinesis Firehose, CloudWatch Logs, S3), resource association (ALB, API Gateway, CloudFront), rate-based rules with aggregate keys (IP, forwarded IP), Challenge and CAPTCHA actions, and account takeover prevention. Emits a READY_TO_DEPLOY checklist. Use when creating a Web ACL, attaching managed rules, configuring rate limiting, enabling WAF logging, associating a Web ACL with an ALB/API Gateway/CloudFront, or configuring CAPTCHA/Challenge actions. Triggers: create WAF Web ACL, WAFv2, managed rule groups, WAF rate limiting, WAF logging, WAF CAPTCHA, WAF Challenge action, bot control, ATP managed rules.

★ 4
jhkchan
jhkchan/aws-agent-skills

wafv2-web-acl-auditor

Audits AWS WAFv2 Web ACL configurations to determine whether the ACL provides effective protection: default-action posture (Allow vs Block), managed-rule-group coverage gaps, rule effectiveness (BLOCK vs COUNT, shadow/bypass rules, stale exclusions), rate-based rule correctness, logging and visibility configuration, and text-transformation bypass vectors. Emits a deterministic verdict per Web ACL. Use when reviewing a WAFv2 Web ACL for security posture, checking managed-rule coverage, validating rate-limit configuration, auditing WAF logging, or hardening a WAF before production deployment. Triggers: WAF, WAFv2, Web ACL, managed rule group, AWSManagedRules, rate-based rule, CAPTCHA, Challenge, OverrideAction, Count mode, forwarded IP, Bot Control, ATP, text transformation, WAF logging, WAF visibility, OWASP, SQLi, XSS, RFI.

★ 4
jhkchan
jhkchan/aws-agent-skills

waf-rule-deployer

Provisions AWS WAFv2 rule sets with production defaults: Web ACL creation (CloudFront scope in us-east-1 vs Regional scope), rule groups (managed vs custom) with capacity planning in WCU (Web ACL Capacity Units, 1500 default per ACL), managed rule groups (AWSManagedRulesCommonRuleSet, AWSManagedRulesSQLiRuleSet, AWSManagedRulesAmazonIpReputationList, AWSManagedRulesLinuxRuleSet, AWSManagedRulesWindowsRuleSet, AWSManagedRulesUnixRuleSet), custom rules (byte-match, regex pattern set, geo-match, rate-based), IP set and regex pattern set resources, rule priority ordering (lowest number evaluated first), action types (allow, block, count, CAPTCHA, Challenge), bot control managed rule group, account takeover. Triggers: create WAF rule, WAF managed rule group, WAF custom rule, WAF byte match, WAF geo match, WAF rate based rule, WAF IP set, WAF regex pattern set, WAF WCU budget, WAF CloudFront ACL, WAF logging Kinesis Firehose, WAF bot control, WAF ATP, WAF label matching, WAF Challenge action, WAF CAPTCHA.

★ 4