ssm-association-operator
Operates AWS Systems Manager (SSM) State Manager associations with production defaults: association creation (document name, targets, schedule), association status (Success/Failed/Pending), compliance reporting, association versioning, output location (S3 bucket with KMS-encrypted SSE), rate control (max-concurrency and max-errors as integer counts, NOT percentages), target via resource tags (dynamic — new instances auto-included) vs instance IDs (static), association automation (apply on cron/rate schedule), association remediation on non-compliance, multi-document association, document-specific parameters, SSM document lifecycle (create/update/default version), and CloudWatch metrics for association execution. Emits an OPERATION_COMPLETED. Triggers: create ssm association, ssm state manager, ssm association schedule, ssm association rate control, ssm association targets, ssm association output s3, ssm association compliance, ssm association version, ssm document parameters, ssm association remediation.
- License 许可
- Apache-2.0
- 最近更新
- 2026-09-24
- Stars
- ★ 4
- Forks
- 0
01 前言 Front Matter
- name
- ssm-association-operator
- license
- Apache-2.0
- metadata
- { "tags": "aws, systems-manager, ssm, state-manager, association, cloudops, operate, compliance, rate-control, ssm-document", "phase": "2", "author": "Jacky Chan — AWS Community Builder", "domain": "aws-cloudops", "family": "Management", "version": "0.1.0", "keywords": "aws, systems manager, ssm, state manager, association, create association, association schedule, association targets, rate control, max-concurrency, max-errors, compliance, association version, output location, ssm document, document lifecycle, remediation, tag targets, cloudops, operate", "task_type": "operate", "complexity": "medium", "entry_point": "false", "skill_class": "capability", "when_to_use": "Invoke when the user wants to create or operate an SSM State Manager association — apply a document (AWS-managed or custom) to targets on a schedule, target instances by tag (dynamic) or instance IDs (static), configure rate control (max-concurrency and max-errors as integer counts), route execution output to an S3 bucket, view association status and compliance, version an association, set up remediation when an association drifts to non-compliant, or run multi-document associations. Do NOT invoke for ad-hoc one-off command runs (use ssm-run-command skills), patch baseline creation (use ssm-patch-baseline-deployer), or session manager troubleshooting (use ssm-session-troubleshooter).", "dependencies": "aws-orchestrator", "requires_llm": "true", "verdict_shape": "OPERATION_COMPLETED | REVIEW_REQUIRED", "lifecycle_status": "active", "supports_pipeline": "true" }
- description
- Operates AWS Systems Manager (SSM) State Manager associations with production defaults: association creation (document name, targets, schedule), association status (Success/Failed/Pending), compliance reporting, association versioning, output location (S3 bucket with KMS-encrypted SSE), rate control (max-concurrency and max-errors as integer counts, NOT percentages), target via resource tags (dynamic — new instances auto-included) vs instance IDs (static), association automation (apply on cron/rate schedule), association remediation on non-compliance, multi-document association, document-specific parameters, SSM document lifecycle (create/update/default version), and CloudWatch metrics for association execution. Emits an OPERATION_COMPLETED. Triggers: create ssm association, ssm state manager, ssm association schedule, ssm association rate control, ssm association targets, ssm association output s3, ssm association compliance, ssm association version, ssm document parameters, ssm association remediation.
- compatibility
- Agent runtime that reads SKILL.md (Claude Code, Cursor, Windsurf, Codex, Gemini). For live deployment: AWS CLI v2 with ssm access (and iam access for role/policy creation, s3 access for output bucket validation). Works with Terraform aws_ssm_association / aws_ssm_document resources and CloudFormation AWS::SSM::Association / AWS::SSM::Document templates.
02 资源文件 19
- 参考文档eval/test-cases.yaml7.1 KB
- 参考文档evals/baselines/compliance-remediation-association.md722 B
- 参考文档evals/baselines/instance-id-target-association.md740 B
- 参考文档evals/baselines/missing-s3-kms-key.md883 B
- 参考文档evals/baselines/multi-document-step-association.md871 B
- 参考文档evals/baselines/tag-target-scheduled-association.md1022 B
- 参考文档evals/eval_config.yaml1.8 KB
- 参考文档evals/evals.json10.8 KB
- 参考文档evals/prompts/compliance-remediation-association.md775 B
- 参考文档evals/prompts/instance-id-target-association.md629 B
- 参考文档evals/prompts/missing-s3-kms-key.md580 B
- 参考文档evals/prompts/multi-document-step-association.md811 B
- 参考文档evals/prompts/tag-target-scheduled-association.md757 B
- 参考文档examples/README.md8.2 KB
- 参考文档references/advanced-patterns.md944 B
- 参考文档references/error-handling.md1.3 KB
- 参考文档references/output-and-compliance.md13.6 KB
- 参考文档references/rate-control-and-targets.md12.9 KB
- 参考文档skill.usf.yaml4.5 KB
03 SKILL.md 正文
SSM Association Operator
An AWS CloudOps agent skill that operates AWS Systems Manager (SSM) State Manager associations with correct defaults. The skill walks the operator through the document/target/schedule model, tag-vs-instanceID targeting, integer rate control, S3 output with KMS-encrypted SSE, compliance reporting, association versioning, and remediation on non-compliance, captures the configuration, explains why each default matters, and emits an OPERATION_COMPLETED checklist with copy-pasteable verification commands.
Activation keywords
create SSM association, SSM State Manager, SSM association schedule, SSM association rate control, SSM association targets, SSM association output S3, SSM association compliance, SSM association version, SSM document parameters, SSM association remediation.
STRICT output contract
When this skill is invoked with an SSM-association-operation request
(create an association, target instances by tag, schedule a document
run, configure rate control, route output to S3, set up remediation,
version an association, or report compliance status), the agent MUST
respond with the OPERATION_COMPLETED checklist defined in the "Output
format" section using the literal all-caps labels SSM_ASSOCIATION:,
VERDICT:, CHECKLIST:, and VERIFICATION_COMMANDS:. Do NOT preface
the checklist with prose, headings, or disclaimers — emit the block as
the first lines of the response. This contract is what assertion-based
evals and downstream operations pipelines rely on; deviating from the
literal labels breaks automation silently.
If any prerequisite is missing or the configuration shows a material
gap (S3 output bucket without KMS, percentage-string rate control,
schedule without apply-at-creation, etc.), the verdict is
REVIEW_REQUIRED with a specific gap citation in the checklist (marked
[!]), and OPERATION_COMPLETED MUST NOT also appear.
Quick navigation
| Section | When to read |
|---|---|
| Prerequisites | Always — verify before operating |
| Step 1 — Document/target/schedule model | Core association model |
| Step 2 — Targets: tag vs instance IDs | Targeting decision |
| Step 3 — Rate control (integer counts, NOT percentages) | Concurrency + error threshold |
| Step 4 — Schedule (cron/rate, apply-at-creation) | When associations run |
| Step 5 — Output location (S3 + KMS-encrypted SSE) | Execution artifacts |
| Step 6 — Association status and compliance | Success/Failed/Pending + compliance |
| Step 7 — Association versioning | Update + default version |
| Step 8 — Remediation on non-compliance | Auto-remediation wiring |
| Step 9 — Multi-document associations | Documents chaining |
| Step 10 — SSM document lifecycle | Create/update/default version |
| Step 11 — CloudWatch metrics | Observability |
| Step 12 — Recent features | Latest |
| NEVER do these things | Review before signing off |
| Output format | The literal checklist template |
| references/rate-control-and-targets.md | Rate control + targeting detail |
| references/output-and-compliance.md | S3 output + compliance detail |
Mindset
One-line takeaway: An SSM State Manager association applies a
document (AWS-managed or custom) to a set of targets on a schedule.
Targets specified by tag are DYNAMIC — new instances matching the tag
are picked up automatically on the next run. Targets specified by
instance ID are STATIC — new instances are NOT picked up until the
association is updated. Rate control uses max-concurrency and
max-errors as INTEGER COUNTS (e.g., 10 or 1), NOT percentages.
Three misconceptions dominate SSM association misoperation:
-
"Rate control accepts percentages as the canonical form." It does, but the integer count form is preferred for deterministic behavior across fleets of varying size. A
10%value on a 3-instance fleet rounds unpredictably;max-concurrency: 1is unambiguous. The skill always emits the integer form and flags percentage strings as REVIEW_REQUIRED. -
"Tag-based targets behave the same as instance-ID targets." They do not. Tag targets are dynamic: any instance that gains the tag (e.g., newly launched by Auto Scaling) is automatically picked up at the next association run. Instance-ID targets are static: the association only ever touches the IDs explicitly listed. For autoscaling fleets, ALWAYS use tag targets.
-
"The S3 output bucket just needs to exist." It needs SSE enabled, and for production that means a customer-managed KMS key. SSM does not enforce SSE, but execution artifacts (stdout/stderr/scripts) can contain secrets; shipping them to an unencrypted bucket is a finding. The skill flags a missing KMS key as REVIEW_REQUIRED.
Configuration dependency graph
SSM association configurations are NOT independent. The document must exist before the association is created. The instance profile must allow SSM before targets can run anything. Output S3 bucket policy must allow the SSM service principal to write before output is delivered. Use this graph to sequence operations.
| Configuration | Hard dependencies | Silent failure | Enables downstream |
|---|---|---|---|
| SSM document | document name exists in account or AWS-managed catalog | custom docs need a default version set before reference | the document body |
| Instance profile | target EC2 instances have AmazonSSMManagedInstanceCore | instances without the role show PingStatus: Inactive and are silently skipped | SSM agent reachability |
| Association (create) | document name; targets; schedule | does NOT run at creation unless --apply-at-creation is set |
the association itself |
| Rate control | integer counts for max-concurrency and max-errors | percentage strings work but are size-dependent and ambiguous | safe rollout pace |
| Output S3 bucket | bucket exists; SSM service principal has s3:PutObject | without KMS key, output is unencrypted at rest; SSM silently drops output if bucket policy missing | execution artifacts |
| Schedule | cron or rate expression | rate(30 minutes) and cron(0 30 * * * *) are NOT interchangeable |
automated runs |
| Compliance | association has executed at least once | non-compliant status only computed AFTER first run; before that = Pending | drift detection |
| Remediation | association exists; document supports idempotent apply | remediation re-applies document on non-compliance; non-idempotent docs cause side effects | auto-remediation |
The integer-rate-control row is the one a baseline model misses.
Many baseline answers emit max-concurrency: "10%" because the API
technically accepts percentage strings. The skill normalizes to
integer counts because that is the deterministic, fleet-size-
independent form. The S3-output-needs-KMS row is the second commonly
missed configuration.
Expert heuristic: integer rate control vs percentage strings
→ Moved to references/rate-control-and-targets.md — integer-vs-percentage decision model and worst-case blast radius.
Expert heuristic: tag targets vs instance-ID targets
→ Moved to references/rate-control-and-targets.md — tag-target vs instance-ID decision model.
Expert heuristic: S3 output bucket encryption
→ Moved to references/output-and-compliance.md — S3 output bucket encryption requirements and silent-finding risk.
Prerequisites (verify before operating)
Before emitting association commands, verify these prerequisites. If any are missing, the verdict is REVIEW_REQUIRED.
| Prerequisite | Why it matters | How to verify |
|---|---|---|
| SSM document exists | Association references a document name | aws ssm describe-document --name <name> |
| Default document version set | Custom documents need a default version | aws ssm describe-document --name <name> --query 'Document.DefaultVersion' |
| Target instances managed by SSM | Targets must have PingStatus: Online | aws ssm describe-instance-information --filters Key=Tag:Environment,Values=[production] |
| Output S3 bucket exists | Output location must be writable | aws s3api head-bucket --bucket <bucket> |
| KMS key for output bucket | SSE-KMS required for production | aws s3api get-bucket-encryption --bucket <bucket> |
| Schedule expression valid | Cron needs seconds field; rate needs units | Validate cron(...) or rate(...) syntax |
Step 1 — Document/target/schedule model
An SSM State Manager association binds three things: a document, a set of targets, and a schedule.
| Component | What it specifies | API field |
|---|---|---|
| Document | What to run (AWS-managed or custom) | --name AWS-ApplyPatchBaseline |
| Targets | Which instances to run on | --targets Key=tag:Environment,Values=production |
| Schedule | When to run | --schedule-expression "rate(30 minutes)" |
| Parameters | Document-specific inputs | --parameters Operation=Install |
AWS-managed documents include AWS-ApplyPatchBaseline,
AWS-GatherSoftwareInventory, AWS-UpdateSSMAgent,
AWS-ConfigureCloudWatchOnEC2Instances. Custom documents are
user-authored YAML/JSON defining a sequence of steps; they need a
default version set before they can be referenced.
Step 2 — Targets: tag vs instance IDs
| Target type | Dynamic? | Use case | Example |
|---|---|---|---|
| Tag | Yes — new tagged instances auto-included | Auto Scaling fleets | Key=tag:Environment,Values=production |
| Instance IDs | No — only listed IDs targeted | Pet servers, one-off | Key=InstanceIds,Values=i-aaa111,i-bbb222 |
| Resource group | Yes — membership rules | Complex grouping | Key=ResourceGroups,Values=<rg-arn> |
For any fleet with churn, ALWAYS use tag targets.
Step 3 — Rate control (integer counts, NOT percentages)
aws ssm create-association \
--name "AWS-ApplyPatchBaseline" \
--targets "Key=tag:Environment,Values=production" \
--schedule-expression "rate(30 minutes)" \
--max-concurrency 10 \
--max-errors 3 \
--parameters "Operation=Install" \
--output-location '{"S3Location":{"OutputS3BucketName":"my-ssm-output","OutputS3KeyPrefix":"ssm-output/","OutputS3Region":"us-east-1"}}' \
--apply-at-creation \
--region us-east-1
Critical: the integer form 10 is unambiguous across fleet sizes.
The percentage form "10%" rounds unpredictably in small fleets. The
skill normalizes to integers and flags percentage strings as
REVIEW_REQUIRED.
Step 4 — Schedule (cron/rate, apply-at-creation)
| Expression | Format | Example | Notes |
|---|---|---|---|
rate |
rate(<value> <units>) |
rate(30 minutes) |
Min: 1 minute; no seconds |
cron |
cron(<sec> <min> <hr> <dom> <mon> <dow> <year>) |
cron(0 30 2 ? * * *) |
6 required fields + optional year; AWS cron (NOT Unix) |
Critical: AWS cron is NOT Unix cron. AWS cron has 6 fields (with optional 7th year), starting with SECONDS. Unix cron has 5 fields starting with minutes. The skill always uses AWS cron syntax.
Apply-at-creation: by default, an association does NOT run when
created. It waits for the next scheduled tick. For first-run
validation, set --apply-at-creation. The skill always sets it.
Step 5 — Output location (S3 + KMS-encrypted SSE)
aws ssm create-association \
--name "AWS-ApplyPatchBaseline" \
--targets "Key=tag:Environment,Values=production" \
--schedule-expression "rate(30 minutes)" \
--output-location '{
"S3Location": {
"OutputS3BucketName": "my-ssm-output",
"OutputS3KeyPrefix": "ssm-output/",
"OutputS3Region": "us-east-1"
}
}' \
--region us-east-1
Verify the bucket has KMS-encrypted SSE:
aws s3api get-bucket-encryption --bucket my-ssm-output --region us-east-1
# Expected: ServerSideEncryptionConfiguration with SSE-KMS + KMSMasterKeyID
Critical: SSM does NOT enforce SSE on the output bucket. An
unencrypted bucket is a silent finding. The skill flags any output
bucket without a customer-managed KMS key as REVIEW_REQUIRED. The S3
bucket policy must allow the SSM service principal to s3:PutObject,
and the KMS key policy must allow SSM to kms:GenerateDataKey.
Step 6 — Association status and compliance
| Status | Meaning | Action |
|---|---|---|
Success |
All targeted instances ran successfully | None |
Failed |
One or more instances errored; rate control may have stopped the run | Inspect per-instance output in S3 |
Pending |
Association created but not yet executed | Wait, or trigger with --apply-at-creation |
TimedOut |
Document execution exceeded timeout | Increase timeout or fix document |
# Association status + execution history
aws ssm describe-association --association-id "<id>" --region us-east-1
aws ssm describe-association-executions --association-id "<id>" --region us-east-1
# Compliance reporting
aws ssm list-compliance-items --filters "Key=ComplianceType,Values=Association" --region us-east-1
aws ssm list-resource-compliance-summaries --filters "Key=ComplianceType,Values=Association" --region us-east-1
Step 7 — Association versioning
Each association update creates a new version (1, 2, 3, ...).
aws ssm update-association \
--association-id "<id>" \
--schedule-expression "rate(15 minutes)" \
--region us-east-1
aws ssm list-association-versions --association-id "<id>" --region us-east-1
Critical: there is no built-in rollback command. To revert, read
the desired version's parameters from list-association-versions and
re-apply them via update-association.
Step 8 — Remediation on non-compliance
When an association drifts to non-compliant, SSM can auto-remediate by re-applying the document.
NON_COMPLIANT detected
→ EventBridge rule pattern:
source: [aws.ssm]
detail-type: [Configuration Compliance State Change]
detail.status: [NON_COMPLIANT]
→ Target: SSM Automation or Lambda → StartAssociationsOnce
→ SSM re-applies the document → compliance flips back
Manual re-run:
aws ssm start-associations-once --association-ids "<id>" --region us-east-1
Critical: remediation only works if the document is IDEMPOTENT. A non-idempotent document (e.g., one that appends to a file) will cause side effects on re-run. Always verify idempotency before wiring remediation.
Step 9 — Multi-document associations
A single association applies ONE document. To apply multiple documents to the same targets, create SEPARATE associations, OR build a single custom step-document that runs multiple steps in sequence:
schemaVersion: '2.2'
name: Custom-ProvisionAndConfigure
mainSteps:
- action: aws:runShellScript
name: installPackage
inputs:
runCommand:
- yum install -y httpd
- action: aws:runShellScript
name: configureService
inputs:
runCommand:
- systemctl enable httpd
- systemctl start httpd
Use a step-document when the steps have ordering dependencies. Use separate associations when the documents are independent.
Step 10 — SSM document lifecycle
# Create
aws ssm create-document --content file://doc.yaml --name "Custom-Harden" \
--document-type Command --document-format YAML --region us-east-1
# Update (creates new version)
aws ssm update-document --content file://doc-v2.yaml --name "Custom-Harden" \
--document-version "1" --region us-east-1
# Set default version (associations use the default version)
aws ssm update-document-default-version --name "Custom-Harden" \
--document-version "2" --region us-east-1
Critical: an association referencing a custom document always uses
the document's DEFAULT VERSION. To pin an association to a specific
version, reference it explicitly: Custom-HardenBaseline:2.
Step 11 — CloudWatch metrics
SSM emits metrics in the AWS/SSM namespace.
| Metric | Description |
|---|---|
AssociationSuccess |
Count of successful executions |
AssociationFailed |
Count of failed executions — alarm on spikes |
AssociationTimeout |
Count of timed-out executions |
AssociationCompliance |
1=compliant, 0=non-compliant |
Step 12 — Recent features
→ Moved to references/advanced-patterns.md — recent AWS features (2023-2026).
NEVER do these things
-
NEVER use percentage strings for rate control.
max-concurrency: "10%"behaves differently across fleet sizes. Use integer counts (max-concurrency: 10,max-errors: 3). -
NEVER use instance-ID targets for autoscaling fleets. Instance IDs are static; new instances are NOT picked up. Use tag targets.
-
NEVER configure an output S3 bucket without customer-managed KMS-encrypted SSE. SSM does not enforce SSE; an unencrypted bucket is a silent finding.
-
NEVER create an association without
--apply-at-creationfor the first run. Otherwise it waits for the next scheduled tick. -
NEVER use Unix cron syntax. AWS cron has 6 fields starting with SECONDS.
cron(0 30 2 ? * * *)is AWS cron for 02:30 daily. -
NEVER wire remediation for a non-idempotent document. Remediation re-applies the document; non-idempotent documents cause side effects on re-run.
-
NEVER assume compliance status is computed before the first run. A newly created association shows "Pending" until it executes at least once.
-
NEVER update a custom document without setting a default version. Associations reference the default version.
-
NEVER assume the S3 output bucket policy is correct without testing. SSM silently drops output if the bucket policy does not allow the SSM service principal to
s3:PutObject. -
NEVER assume the association will run immediately after creation. By default, the first run happens at the next scheduled execution.
Output format
SSM_ASSOCIATION: <association-name> (<association-id>)
VERDICT: OPERATION_COMPLETED | REVIEW_REQUIRED
CHECKLIST:
[✓|!] Document: <document-name> (version <v|default>)
[✓|!] Targets: <tag-key>=<values> (dynamic) | InstanceIds=<ids> (static)
[✓|!] Schedule: <cron|rate expression>
[✓|!] Apply-at-creation: ENABLED | DISABLED
[✓|!] Rate control: max-concurrency=<N>, max-errors=<N> (integer form)
[✓|!] Output S3: s3://<bucket>/<prefix> (SSE-KMS) | UNENCRYPTED [!]
[✓|!] Parameters: <key=value list>
[✓|!] Compliance: COMPLIANT | NON_COMPLIANT | PENDING (first run)
[✓|!] Status: Success | Failed | Pending
[✓|!] Association version: <v>
[✓|!] CloudWatch alarm: <alarm-name or "none">
[✓|!] Tags: <key=value list>
VERIFICATION_COMMANDS:
aws ssm describe-association --association-id <association-id> --region <region>
aws ssm describe-association-executions --association-id <association-id> --region <region>
aws ssm list-compliance-items --filters Key=ComplianceType,Values=Association --region <region>
aws s3api get-bucket-encryption --bucket <output-bucket> --region <region>
Worked example — tag-targeted scheduled association with integer rate control and KMS-encrypted S3 output
SSM_ASSOCIATION: PatchProductionFleet (a-1a2b3c4d5e6f7g8h9)
VERDICT: OPERATION_COMPLETED
CHECKLIST:
[✓] Document: AWS-ApplyPatchBaseline (default version)
[✓] Targets: tag:Environment=production (dynamic)
[✓] Schedule: rate(30 minutes)
[✓] Apply-at-creation: ENABLED
[✓] Rate control: max-concurrency=10, max-errors=3 (integer form)
[✓] Output S3: s3://my-ssm-output/ssm-output/ (SSE-KMS, key arn:aws:kms:us-east-1:123456789012:key/abc123)
[✓] Parameters: Operation=Install, SnapshotId=latest
[✓] Compliance: COMPLIANT
[✓] Status: Success
[✓] Association version: 3
[✓] CloudWatch alarm: ssm-patch-failures-prod
[✓] Tags: Environment=production, Owner=cloudops
VERIFICATION_COMMANDS:
aws ssm describe-association --association-id a-1a2b3c4d5e6f7g8h9 --region us-east-1
aws ssm describe-association-executions --association-id a-1a2b3c4d5e6f7g8h9 --region us-east-1
aws ssm list-compliance-items --filters Key=ComplianceType,Values=Association --region us-east-1
aws s3api get-bucket-encryption --bucket my-ssm-output --region us-east-1
Error handling
→ Moved to references/error-handling.md — six failure modes: zero targets, stuck Pending, output not delivered, Failed status, NON_COMPLIANT, rate-control stop.
References (load on demand)
- references/rate-control-and-targets.md — pre-existing; extended with the integer-rate-control and tag-vs-instance-ID expert heuristics moved from SKILL.md.
- references/output-and-compliance.md — pre-existing; extended with the S3 output bucket encryption expert heuristic moved from SKILL.md.
- references/advanced-patterns.md — recent AWS features (2023-2026): association-level KMS, target-locations, calendar schedules, EventBridge triggers, compliance dashboard.
- references/error-handling.md — error-handling deep dive: zero targets, Pending stall, S3 output drop, Failed runs, NON_COMPLIANT after success, rate-control stop.
Domain
AWS CloudOps / AWS Systems Manager (SSM) State Manager Association Operations & Compliance Management.
AWS documentation
- SSM State Manager — https://docs.aws.amazon.com/systems-manager/latest/userguide/systems-manager-state.html
- Creating associations — https://docs.aws.amazon.com/systems-manager/latest/userguide/sysman-state-assoc.html
- Targeting instances — https://docs.aws.amazon.com/systems-manager/latest/userguide/sysman-state-target.html
- Rate control — https://docs.aws.amazon.com/systems-manager/latest/userguide/sysman-state-rate-control.html
- Cron schedules — https://docs.aws.amazon.com/systems-manager/latest/userguide/reference-cron-and-rate-expressions.html
- S3 output — https://docs.aws.amazon.com/systems-manager/latest/userguide/sysman-rc-output.html
- Compliance — https://docs.aws.amazon.com/systems-manager/latest/userguide/sysman-compliance.html
- Association versions — https://docs.aws.amazon.com/systems-manager/latest/userguide/sysman-state-association-versions.html
- SSM documents — https://docs.aws.amazon.com/systems-manager/latest/userguide/documents.html
- CloudWatch metrics for SSM — https://docs.aws.amazon.com/systems-manager/latest/userguide/monitoring-cloudwatch-metrics.html
04 同仓库更多 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.
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.
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.
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.
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.
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.