AWS Incident Response Playbook Samples

June 26, 2026 · View on GitHub

These playbooks are provided as templates for organizations building incident response capability on AWS. They should be customized to suit your specific needs, risks, available tools, and work processes. These guides are not official AWS documentation and are provided as-is.

All playbooks are aligned to the NIST SP 800-61 Revision 3: Incident Response Recommendations and Considerations for Cybersecurity Risk Management framework and mapped to the CSF 2.0 Community Profile.

What's new in this version: This repository was significantly refreshed in 2026. Playbooks have been rewritten to align with NIST SP 800-61r3 (previously r2), reference current AWS services (including the AWS Security Incident Response service, IAM Access Analyzer, Amazon Macie, and GuardDuty Runtime Monitoring), and follow a standardized template structure. AI-assisted playbook variants and automation patterns have been added.


Getting Started

ResourcePurpose
Triage GuideQuickly assess incident severity (P1–P4) and determine response urgency
Playbook TemplateStandard structure for all playbooks — use when creating new ones
Regulatory ContextNotification obligations by incident type and regulation
ContributingHow to contribute new playbooks or improvements

Playbooks

Identity & Access Scenarios

PlaybookDescription
Credential CompromiseIAM access key or console credential theft and abuse
STS Token AbuseAssumeRole chain attacks, cross-account pivoting, IMDS credential theft
Identity Center CompromiseAWS SSO permission set abuse, identity store manipulation
Federated Access AbuseBEC or IdP compromise (Okta, Azure AD) leading to AWS access
Insider ThreatAnomalous authorized user behavior — technical detection and containment

Compute & Infrastructure Scenarios

PlaybookDescription
EC2 CompromiseInstance-level compromise — C2, lateral movement, IMDS abuse
Container/EKS CompromisePod escape, malicious images, IRSA abuse, Kubernetes RBAC manipulation
CI/CD CompromisePipeline compromise, dependency poisoning, build artifact tampering
CryptominingUnauthorized compute resource abuse for cryptocurrency mining
RansomwareCloud-native ransomware — EBS encryption, S3 deletion, KMS abuse
Denial of ServiceDDoS and application-layer attacks

Data & Application Scenarios

PlaybookDescription
Data AccessUnauthorized access to data stores (S3, DynamoDB, Secrets Manager)
S3 Data ExfiltrationBulk S3 extraction — replication abuse, presigned URLs, batch operations
Root Account TakeoverRoot credential compromise — always P1
Personal Data BreachRegulatory notification workflow when personal data is involved

AI / Agentic Workload Scenarios

Incident response for Amazon Bedrock AgentCore — autonomous AI agents whose blast radius expands with every tool call. These scenarios cover the AgentCore-specific attack surface (Cedar authorization, Token Vault credentials, sandbox tools, Memory, observability) that traditional compute/identity playbooks do not address.

PlaybookDescription
AgentCore Identity & Credential CompromiseStolen Cognito JWT, machine-client secret, workload-identity session, or Token Vault / OAuth2 / API-key credential-provider theft
AgentCore Agent IntegrityPrompt injection, memory poisoning, poisoned Runtime artifact (S3 ZIP / ECR image), Registry supply-chain compromise
AgentCore Authorization BypassCedar Policy Engine flipped ENFORCE → LOG_ONLY, rogue Gateway target, cross-account resource-based policy
AgentCore Tool AbuseCode Interpreter SSRF / exfiltration, Browser SSRF / saved-profile persistence, network-mode drift to PUBLIC
AgentCore Observability TamperingCloudTrail / Log Group / X-Ray / KMS tampering that blinds the investigator; Evaluations-role trace exfiltration

GuardDuty Finding Quick-Response Guides

These are finding-specific quick-response guides (5-minute triage) that route to the full lifecycle playbooks above.

CategoryFindings Covered
EC2 FindingsBackdoor, Trojan, CryptoCurrency, UnauthorizedAccess, Recon, Impact
S3 FindingsExfiltration, Discovery, Impact, Policy, UnauthorizedAccess
IAM FindingsCredentialAccess, Discovery, InitialAccess, Persistence, PrivilegeEscalation
EKS & Runtime FindingsExecution, PrivilegeEscalation, Persistence, Discovery, DefenseEvasion

AI-Assisted Playbooks

The ai-playbooks/ directory contains vendor-agnostic, AI-readable versions of these playbooks. They work with any AI coding assistant or LLM — Kiro, Claude Code, Cursor, Windsurf, GitHub Copilot, or any tool that can consume markdown as context.

The AI playbooks are designed with human-in-the-loop safeguards: the AI guides and recommends, but always waits for explicit human confirmation before executing any action that modifies the environment.

See the AI Playbooks README for setup instructions and architecture details.


Automation Patterns

The automation-patterns/ directory contains reference examples showing how to wire AWS services for IR automation. These are starting points, not production-ready IaC.

PatternDescription
EventBridge Rules10 event patterns for common IR triggers (GuardDuty, root activity, CloudTrail tampering)
Step Functions WorkflowReference state machine for IR orchestration with human approval gates
Security Hub Custom ActionsHuman-in-the-loop automation — click a button to isolate, revoke, or preserve

For production-ready implementations, see aws-samples/aws-security-incident-response-integrations.


Getting Help from AWS

All AWS customers can request assistance from the AWS Customer Incident Response Team (CIRT) through a support case, regardless of support plan level. You do not need any specific service subscription to get help during a security incident.

Additionally, the AWS Security Incident Response service provides automated triage, case management, and proactive engagement capabilities for customers who enable it.


ResourceDescription
AWS Security Incident Response Guide (2023)Comprehensive guide to IR on AWS
NIST SP 800-61r3Incident Response Recommendations and Considerations
AWS Customer Playbook FrameworkAdditional playbook templates with multi-language support
AWS Security IR IntegrationsSample automation integrations for IR workflows
AWS CIRT WorkshopsHands-on IR workshops
AWS Threat Detection & Response WorkshopWorkshop for detection and response
AWS Security Reference ArchitecturePrescriptive security architecture guidance

Usage

These playbooks are written in markdown to facilitate editing and consumption into a variety of systems. They should be tested (for example, in Game Days) prior to deployment and all responders should be familiar with the actions required.

Some incident response steps may incur costs in your AWS account(s). Customizing and testing these scenarios will help you determine potential cost impact.


Security

See CONTRIBUTING for more information.


License Summary

The documentation is made available under the Creative Commons Attribution-ShareAlike 4.0 International License. See the LICENSE file.

The sample code within this documentation is made available under the MIT-0 license. See the LICENSE-SAMPLECODE file.