{
 "name": "The Agent Boundary Checklist",
 "edition": "2026-11-10",
 "read_date": "2026-11-10",
 "source": "https://aiagentaccesscontrol.com/datasets/",
 "license": "CC BY 4.0",
 "cite": "Cite this line: 'AI Agent Access Control, The Agent Access Map, edition 10 Nov 2026, https://aiagentaccesscontrol.com/datasets/'",
 "description": "30 checks with the mechanism in AWS, Google Cloud and Azure and the documentation source.",
 "rows": [
  {
   "id": "ab-01",
   "section": "identity",
   "check": "Give each agent its own role or service account, not a shared one",
   "why": "A shared identity means one agent's access is every agent's access, and logs cannot say which agent acted. A separate identity lets you grant, review and revoke access agent by agent.",
   "how_aws": "Use one Lambda execution role per function, or one EKS Pod Identity association per service account, so each agent maps to its own IAM role.",
   "how_gcp": "Create dedicated service accounts for each application, or use a per-agent agent identity on Agent Runtime.",
   "how_azure": "Use the per-agent Entra agent identity that Foundry creates, or a system-assigned managed identity that is tied to one container app.",
   "sources": "https://docs.aws.amazon.com/lambda/latest/dg/lambda-intro-execution-role.html https://docs.aws.amazon.com/eks/latest/userguide/pod-identities.html https://cloud.google.com/iam/docs/best-practices-service-accounts https://docs.cloud.google.com/gemini-enterprise-agent-platform/scale/runtime/agent-identity https://learn.microsoft.com/en-us/azure/ai-foundry/agents/concepts/agent-identity https://learn.microsoft.com/en-us/azure/container-apps/managed-identity",
   "read_date": "2026-11-10"
  },
  {
   "id": "ab-02",
   "section": "identity",
   "check": "Use short-lived role credentials and no long-lived access keys for the agent",
   "why": "A key stored in code, a config file or a container image stays valid until someone rotates it, so a leak gives an attacker lasting access. Temporary credentials from a role expire on their own.",
   "how_aws": "Have the workload use temporary credentials through an IAM role rather than IAM user access keys.",
   "how_gcp": "Use workload identity or an attached service account instead of service account keys, which Google describes as powerful credentials that can present a security risk.",
   "how_azure": "Use a managed identity; Microsoft documents that you do not manage its credentials.",
   "sources": "https://docs.aws.amazon.com/IAM/latest/UserGuide/best-practices.html https://cloud.google.com/iam/docs/workload-identity-federation https://learn.microsoft.com/en-us/entra/identity/managed-identities-azure-resources/overview",
   "read_date": "2026-11-10"
  },
  {
   "id": "ab-03",
   "section": "identity",
   "check": "Limit who can pass, attach or impersonate the agent's role",
   "why": "Anyone who can attach the agent's role to a new resource can borrow its access without ever being granted it. Limiting that right keeps the agent's permissions from spreading.",
   "how_aws": "Grant iam:PassRole only for the specific roles a principal needs to pass to a service.",
   "how_gcp": "Grant the Service Account User role only to principals that need to attach the service account to a resource.",
   "how_azure": "none found on 2026-11-10",
   "sources": "https://docs.aws.amazon.com/IAM/latest/UserGuide/id_roles_use_passrole.html https://cloud.google.com/iam/docs/service-account-permissions",
   "read_date": "2026-11-10"
  },
  {
   "id": "ab-04",
   "section": "identity",
   "check": "Scope the role's trust to the specific agent, service account or workload",
   "why": "A role that any workload in an account or cluster can assume is a role every compromised workload can use. Tying trust to one agent keeps a break-in elsewhere from reaching it.",
   "how_aws": "Use trust policy conditions (for Pod Identity, session tags for service account, namespace and cluster; for services, aws:SourceArn and aws:SourceAccount).",
   "how_gcp": "In IAM policies, grant access to the principal that corresponds to one Kubernetes workload, not to the whole cluster or project.",
   "how_azure": "Configure workload identity federation so one managed identity trusts one Kubernetes workload.",
   "sources": "https://docs.aws.amazon.com/eks/latest/userguide/pod-id-role.html https://docs.aws.amazon.com/IAM/latest/UserGuide/confused-deputy.html https://cloud.google.com/kubernetes-engine/docs/concepts/workload-identity https://learn.microsoft.com/en-us/entra/workload-id/workload-identity-federation",
   "read_date": "2026-11-10"
  },
  {
   "id": "ab-05",
   "section": "identity",
   "check": "Do not run the agent as a default or pre-granted account",
   "why": "Default accounts are created for convenience and can carry broad roles that nobody chose for the agent. An agent that inherits one gets access no one reviewed.",
   "how_aws": "Replace the console-created default role with a role you wrote; AWS documents the Lambda console default as an execution role with minimal permissions (basic logging).",
   "how_gcp": "Do not rely on the Compute Engine default service account, which can receive the Editor role; enforce the iam.automaticIamGrantsForDefaultServiceAccounts constraint.",
   "how_azure": "none found on 2026-11-10",
   "sources": "https://docs.aws.amazon.com/lambda/latest/dg/lambda-intro-execution-role.html https://cloud.google.com/iam/docs/best-practices-service-accounts https://cloud.google.com/run/docs/securing/service-identity",
   "read_date": "2026-11-10"
  },
  {
   "id": "ab-06",
   "section": "permissions",
   "check": "Write the agent's permission policy from what its tools actually call",
   "why": "Broad starter policies make every tool call possible, including ones nobody planned. A policy limited to the real calls limits what a wrong or hijacked call can do.",
   "how_aws": "Start from the documented minimum policy and narrow it; AWS says to use broad deployment permissions only for development and testing.",
   "how_gcp": "Grant the agent identity only the roles its code needs; Google describes agent identity as enabling a least-privilege approach.",
   "how_azure": "Grant the least privilege needed and avoid broader roles at broader scopes.",
   "sources": "https://docs.aws.amazon.com/bedrock-agentcore/latest/devguide/runtime-permissions.html https://docs.cloud.google.com/gemini-enterprise-agent-platform/scale/runtime/agent-identity https://learn.microsoft.com/en-us/azure/role-based-access-control/best-practices",
   "read_date": "2026-11-10"
  },
  {
   "id": "ab-07",
   "section": "permissions",
   "check": "Set an organization-level ceiling the agent's role cannot exceed",
   "why": "A role policy can be edited by whoever manages that account, so it is a weak limit by itself. A ceiling set higher in the hierarchy holds even if the role's own policy is widened.",
   "how_aws": "Apply SCPs, which specify the maximum available permissions for IAM roles in the organization, and permissions boundaries on the agent's role.",
   "how_gcp": "Use Principal Access Boundary policies to define the resources the agent's principal may access.",
   "how_azure": "Use deny assignments, which block listed actions even when a role assignment grants them (customer-managed deny assignments are documented as private preview), and apply governance conditions at management group scope.",
   "sources": "https://docs.aws.amazon.com/organizations/latest/userguide/orgs_manage_policies_scps.html https://docs.aws.amazon.com/IAM/latest/UserGuide/access_policies_boundaries.html https://cloud.google.com/iam/docs/principal-access-boundary-policies https://learn.microsoft.com/en-us/azure/role-based-access-control/deny-assignments https://learn.microsoft.com/en-us/azure/governance/management-groups/overview",
   "read_date": "2026-11-10"
  },
  {
   "id": "ab-08",
   "section": "permissions",
   "check": "Explicitly deny the agent IAM changes and destructive actions",
   "why": "An agent that can edit roles or delete data can widen its own access or cause damage no review would have approved. An explicit deny wins over any allow added later by mistake.",
   "how_aws": "Use an SCP or permissions boundary whose maximum excludes IAM write actions and destructive calls on protected resources.",
   "how_gcp": "Use an IAM deny policy that targets agent identities; Google's example denies all agents in an organization from creating roles and deleting buckets.",
   "how_azure": "Use deny assignments to block specific actions even if a role assignment grants them.",
   "sources": "https://docs.aws.amazon.com/IAM/latest/UserGuide/access_policies_boundaries.html https://cloud.google.com/iam/docs/deny-overview https://docs.cloud.google.com/gemini-enterprise-agent-platform/scale/runtime/agent-identity https://learn.microsoft.com/en-us/azure/role-based-access-control/deny-assignments",
   "read_date": "2026-11-10"
  },
  {
   "id": "ab-09",
   "section": "permissions",
   "check": "Compare the agent's granted permissions with what it used and remove the rest",
   "why": "Permissions granted during a pilot tend to stay long after the agent stops using them. Each unused permission is access an attacker could use for nothing the agent needs.",
   "how_aws": "Create an IAM Access Analyzer unused access analyzer to find unused access.",
   "how_gcp": "Use IAM role recommendations to identify and remove excess permissions.",
   "how_azure": "Run recurring Microsoft Entra access reviews of role assignments.",
   "sources": "https://docs.aws.amazon.com/IAM/latest/UserGuide/what-is-access-analyzer.html https://cloud.google.com/iam/docs/recommender-overview https://learn.microsoft.com/en-us/entra/id-governance/access-reviews-overview",
   "read_date": "2026-11-10"
  },
  {
   "id": "ab-10",
   "section": "permissions",
   "check": "Grant roles at the narrowest scope that works",
   "why": "A role granted on a whole account, project or subscription covers every resource in it, including ones created later. Granting on specific resources keeps new resources out of the agent's reach by default.",
   "how_aws": "Name specific resource ARNs in the policy; AWS's Bedrock guidance says to replace the wildcard with specific agent IDs once they exist.",
   "how_gcp": "Add IAM Conditions so a role binding applies only when conditions are met.",
   "how_azure": "Assign the role at resource or resource group scope instead of subscription or management group scope.",
   "sources": "https://docs.aws.amazon.com/bedrock/latest/userguide/agents-permissions.html https://cloud.google.com/iam/docs/conditions-overview https://learn.microsoft.com/en-us/azure/role-based-access-control/overview",
   "read_date": "2026-11-10"
  },
  {
   "id": "ab-11",
   "section": "data",
   "check": "Keep the encryption keys for agent data under your control and limit who can use them",
   "why": "If the agent's role can use a key, it can read whatever that key protects. Customer-managed keys let you cut off access by changing the key policy or disabling the key.",
   "how_aws": "Use a KMS key policy, which AWS calls the primary way to control access to a KMS key.",
   "how_gcp": "Use Cloud KMS customer-managed encryption keys (CMEK) and grant key access only to the identities that need it.",
   "how_azure": "Use customer-managed keys for Azure Storage held in an Azure key store.",
   "sources": "https://docs.aws.amazon.com/kms/latest/developerguide/key-policies.html https://cloud.google.com/kms/docs/cmek https://learn.microsoft.com/en-us/azure/storage/common/customer-managed-keys-overview",
   "read_date": "2026-11-10"
  },
  {
   "id": "ab-12",
   "section": "data",
   "check": "Block public access on storage the agent can write to",
   "why": "An agent that can change a bucket setting, or write to a bucket that already allows public reads, can expose data by accident. Account-level public access blocks stop that even when a policy says otherwise.",
   "how_aws": "Turn on S3 Block Public Access settings.",
   "how_gcp": "Enable public access prevention on buckets or through the organization policy constraint.",
   "how_azure": "Disable anonymous access on storage accounts; Microsoft recommends this.",
   "sources": "https://docs.aws.amazon.com/AmazonS3/latest/userguide/access-control-block-public-access.html https://cloud.google.com/storage/docs/public-access-prevention https://learn.microsoft.com/en-us/azure/storage/blobs/anonymous-read-access-prevent",
   "read_date": "2026-11-10"
  },
  {
   "id": "ab-13",
   "section": "data",
   "check": "Put data stores the agent reads inside a perimeter that blocks copying data out",
   "why": "An agent allowed to read data can usually also send it somewhere else unless something stops the transfer. A perimeter turns an allowed read into a read that cannot leave.",
   "how_aws": "Build a data perimeter with identity, resource and network controls, including VPC endpoint policies.",
   "how_gcp": "Use VPC Service Controls, which Google documents as protecting against the risk of data exfiltration.",
   "how_azure": "Use virtual network service endpoint policies to allow data exfiltration only to specific Azure Storage accounts.",
   "sources": "https://docs.aws.amazon.com/whitepapers/latest/building-a-data-perimeter-on-aws/building-a-data-perimeter-on-aws.html https://docs.aws.amazon.com/vpc/latest/privatelink/vpc-endpoints-access.html https://cloud.google.com/vpc-service-controls/docs/overview https://learn.microsoft.com/en-us/azure/virtual-network/virtual-network-service-endpoint-policies-overview",
   "read_date": "2026-11-10"
  },
  {
   "id": "ab-14",
   "section": "data",
   "check": "Limit data access with tags, attributes or conditions, not only resource lists",
   "why": "Resource lists go stale as new data stores appear. Conditions on tags or attributes keep access tied to what the data is, so a new store does not become reachable by default.",
   "how_aws": "Use attribute-based access control with tags on IAM entities and resources.",
   "how_gcp": "Use IAM Conditions on role bindings.",
   "how_azure": "Use Azure role assignment conditions (Azure ABAC) that filter a role's permissions by attributes.",
   "sources": "https://docs.aws.amazon.com/IAM/latest/UserGuide/introduction_attribute-based-access-control.html https://cloud.google.com/iam/docs/conditions-overview https://learn.microsoft.com/en-us/azure/role-based-access-control/conditions-overview",
   "read_date": "2026-11-10"
  },
  {
   "id": "ab-15",
   "section": "network",
   "check": "Run agent compute on private networks with no direct path to the internet",
   "why": "Compute with a direct internet path can send data out or pull in instructions from anywhere. A private network makes every outside path an explicit choice.",
   "how_aws": "Attach functions to private subnets; AWS says a public subnet does not give a Lambda function internet access.",
   "how_gcp": "Use a Private Service Connect interface so Agent Runtime traffic stays inside Google's network and never crosses the public internet.",
   "how_azure": "Use the Foundry Standard Setup with private networking, which Microsoft documents as having no public egress.",
   "sources": "https://docs.aws.amazon.com/lambda/latest/dg/configuration-vpc.html https://docs.cloud.google.com/gemini-enterprise-agent-platform/scale/runtime/private-service-connect-interface https://learn.microsoft.com/en-us/azure/ai-foundry/agents/how-to/virtual-networks",
   "read_date": "2026-11-10"
  },
  {
   "id": "ab-16",
   "section": "network",
   "check": "Allow outbound traffic only to the destinations the agent needs",
   "why": "Open outbound access lets a hijacked agent reach any server to send data or fetch tools. A short allow list makes unexpected destinations fail and show up in logs.",
   "how_aws": "Use security group rules and AWS Network Firewall to filter traffic at the perimeter of the VPC.",
   "how_gcp": "Use VPC firewall rules to allow or deny connections, including egress.",
   "how_azure": "Use network security groups to allow or deny outbound traffic, and Azure Firewall for central filtering.",
   "sources": "https://docs.aws.amazon.com/vpc/latest/userguide/vpc-security-groups.html https://docs.aws.amazon.com/network-firewall/latest/developerguide/what-is-aws-network-firewall.html https://cloud.google.com/firewall/docs/firewalls https://learn.microsoft.com/en-us/azure/virtual-network/network-security-groups-overview https://learn.microsoft.com/en-us/azure/firewall/overview",
   "read_date": "2026-11-10"
  },
  {
   "id": "ab-17",
   "section": "network",
   "check": "Reach cloud APIs and data stores through private endpoints",
   "why": "Calls to cloud services over public addresses can be made from anywhere that holds valid credentials. Private endpoints keep the path inside your network, so stolen credentials are harder to use from outside.",
   "how_aws": "Create VPC interface endpoints with AWS PrivateLink.",
   "how_gcp": "Use Private Google Access so VMs with only internal IP addresses can reach Google APIs.",
   "how_azure": "Use private endpoints, which put a service on a private IP address from your virtual network.",
   "sources": "https://docs.aws.amazon.com/bedrock/latest/userguide/usingVPC.html https://cloud.google.com/vpc/docs/private-google-access https://learn.microsoft.com/en-us/azure/private-link/private-endpoint-overview",
   "read_date": "2026-11-10"
  },
  {
   "id": "ab-18",
   "section": "network",
   "check": "Keep the agent's code from reading node, instance or other workloads' credentials",
   "why": "If agent code can read the host's metadata service, it can take the host's credentials, which are often broader than the agent's own. Isolating credentials keeps one compromised agent from becoming a compromised node.",
   "how_aws": "Restrict access to the EC2 Instance Metadata Service for pods so containers can only retrieve credentials for their own role.",
   "how_gcp": "Use Workload Identity Federation for GKE, which protects the metadata that metadata concealment covered.",
   "how_azure": "Run hosted agents in per-session VM-isolated sandboxes, which Microsoft documents as holding no blueprint credential.",
   "sources": "https://docs.aws.amazon.com/eks/latest/userguide/pod-identities.html https://cloud.google.com/kubernetes-engine/docs/concepts/workload-identity https://learn.microsoft.com/en-us/azure/ai-foundry/agents/concepts/agent-identity",
   "read_date": "2026-11-10"
  },
  {
   "id": "ab-19",
   "section": "tools",
   "check": "Give each tool function its own least-privilege role instead of the agent's role",
   "why": "A tool that runs with the agent's full role can do everything the agent can, even if the tool needs one action. A narrow role per tool limits damage when one tool is abused or buggy.",
   "how_aws": "Action group Lambda functions run under their own execution role, and Bedrock reaches them through a resource-based policy.",
   "how_gcp": "Create dedicated service accounts for each application, so each tool service can have its own account.",
   "how_azure": "Use user-assigned managed identities per workload and control which identities are available to the container app.",
   "sources": "https://docs.aws.amazon.com/lambda/latest/dg/lambda-intro-execution-role.html https://docs.aws.amazon.com/bedrock/latest/userguide/agents-lambda.html https://cloud.google.com/iam/docs/best-practices-service-accounts https://learn.microsoft.com/en-us/azure/container-apps/managed-identity",
   "read_date": "2026-11-10"
  },
  {
   "id": "ab-20",
   "section": "tools",
   "check": "Restrict which principals can invoke each tool endpoint",
   "why": "A tool endpoint that accepts calls from any caller can be used without the agent. Limiting invokers means only the intended agent can trigger the action.",
   "how_aws": "Attach a resource-based policy on the tool's Lambda function that allows only the Bedrock service principal for one agent ARN.",
   "how_gcp": "Grant the Cloud Run Invoker role only to the agent's service account.",
   "how_azure": "Assign Foundry roles at agent scope to grant endpoint access to one agent without granting it for all agents in the project.",
   "sources": "https://docs.aws.amazon.com/bedrock/latest/userguide/agents-permissions.html https://cloud.google.com/run/docs/securing/managing-access https://learn.microsoft.com/en-us/azure/ai-foundry/concepts/rbac-foundry",
   "read_date": "2026-11-10"
  },
  {
   "id": "ab-21",
   "section": "tools",
   "check": "Cap how much the agent can run and spend",
   "why": "A looping or hijacked agent can start thousands of runs and a large bill before anyone notices. Limits on concurrency and budget alerts shrink the damage window.",
   "how_aws": "Set Lambda reserved concurrency and AWS Budgets alerts.",
   "how_gcp": "Set a maximum number of Cloud Run instances and Cloud Billing budget alerts.",
   "how_azure": "Set maximum replicas on the container app and Cost Management budgets; Microsoft notes that budget thresholds trigger notifications but do not stop consumption.",
   "sources": "https://docs.aws.amazon.com/lambda/latest/dg/configuration-concurrency.html https://docs.aws.amazon.com/cost-management/latest/userguide/budgets-managing-costs.html https://cloud.google.com/run/docs/configuring/max-instances https://cloud.google.com/billing/docs/how-to/budgets https://learn.microsoft.com/en-us/azure/container-apps/scale-app https://learn.microsoft.com/en-us/azure/cost-management-billing/costs/tutorial-acm-create-budgets",
   "read_date": "2026-11-10"
  },
  {
   "id": "ab-22",
   "section": "tools",
   "check": "Decide for each tool whether the action runs with the agent's authority or the signed-in user's",
   "why": "If a tool runs with the agent's authority, every user can reach whatever the agent can reach. If it runs with the user's authority, the agent can only do what that user may.",
   "how_aws": "AgentCore Gateway lets each target use the gateway role, the caller's credentials, or a token exchange that carries the user and agent identities.",
   "how_gcp": "Google describes separate user, agent and delegated identities; delegated calls are governed by both the agent's policies and the user's permissions.",
   "how_azure": "Foundry documents application-only access (agent's own authority) and delegated access (on behalf of the signed-in user, in preview).",
   "sources": "https://docs.aws.amazon.com/bedrock-agentcore/latest/devguide/gateway-outbound-auth.html https://docs.cloud.google.com/gemini-enterprise-agent-platform/govern/identity-delegation-architecture https://learn.microsoft.com/en-us/azure/ai-foundry/agents/concepts/agent-identity https://learn.microsoft.com/en-us/azure/ai-foundry/agents/concepts/agent-identity",
   "read_date": "2026-11-10"
  },
  {
   "id": "ab-23",
   "section": "change-control",
   "check": "Run agents in their own account, project or subscription",
   "why": "Agents placed beside production workloads share their blast radius, quotas and admin rights. A dedicated container lets you apply stricter policy and limits and remove it cleanly.",
   "how_aws": "Use separate AWS accounts to isolate workloads.",
   "how_gcp": "Use a separate project, grouped under folders in the resource hierarchy.",
   "how_azure": "Use a separate subscription, grouped under management groups.",
   "sources": "https://docs.aws.amazon.com/whitepapers/latest/organizing-your-aws-environment/organizing-your-aws-environment.html https://cloud.google.com/resource-manager/docs/cloud-platform-resource-hierarchy https://learn.microsoft.com/en-us/azure/cloud-adoption-framework/ready/landing-zone/design-area/resource-org-subscriptions",
   "read_date": "2026-11-10"
  },
  {
   "id": "ab-24",
   "section": "change-control",
   "check": "Tag every agent resource with an owner and environment, and enforce the tags",
   "why": "Untagged agent resources cannot be traced to an owner when something goes wrong. Tags also let policy and cost controls target agent resources specifically.",
   "how_aws": "Attach tags to IAM entities and resources; AWS documents tags as the attributes for attribute-based access control.",
   "how_gcp": "Attach tags, which are key-value pairs, to resources.",
   "how_azure": "Use Azure Policy to govern tags.",
   "sources": "https://docs.aws.amazon.com/IAM/latest/UserGuide/introduction_attribute-based-access-control.html https://cloud.google.com/resource-manager/docs/tags/tags-overview https://learn.microsoft.com/en-us/azure/governance/policy/tutorials/govern-tags",
   "read_date": "2026-11-10"
  },
  {
   "id": "ab-25",
   "section": "change-control",
   "check": "Test permission and policy changes in a dry run before enforcing them",
   "why": "A new deny or boundary can break an agent's tools in ways nobody saw coming, and a mistaken allow can open access without anyone noticing. Testing first shows the effect before it reaches production.",
   "how_aws": "Use the IAM policy simulator to test identity-based policies.",
   "how_gcp": "Use organization policies in dry-run mode to monitor how a policy would behave.",
   "how_azure": "Assign Azure Policy with enforcementMode DoNotEnforce.",
   "sources": "https://docs.aws.amazon.com/IAM/latest/UserGuide/access_policies_testing-policies.html https://cloud.google.com/resource-manager/docs/organization-policy/dry-run-policy https://learn.microsoft.com/en-us/azure/governance/policy/concepts/assignment-structure",
   "read_date": "2026-11-10"
  },
  {
   "id": "ab-26",
   "section": "change-control",
   "check": "Remove the agent's identity and access when the agent is retired",
   "why": "Leftover identities from retired agents keep their permissions and nobody is watching them. Tying the identity to the agent's lifecycle removes the access with the agent.",
   "how_aws": "none found on 2026-11-10",
   "how_gcp": "Use agent identity, which Google documents as tied to the lifecycle of the agent.",
   "how_azure": "Use a system-assigned managed identity, which is deleted when the container app is deleted.",
   "sources": "https://docs.cloud.google.com/gemini-enterprise-agent-platform/scale/runtime/agent-identity https://learn.microsoft.com/en-us/azure/container-apps/managed-identity",
   "read_date": "2026-11-10"
  },
  {
   "id": "ab-27",
   "section": "logging",
   "check": "Turn on organization-wide audit logging of control-plane actions",
   "why": "Without audit logs there is no record of which role or setting the agent changed, or who changed the agent's access. A log kept outside the agent's reach is the base for every investigation.",
   "how_aws": "Create an organization trail in CloudTrail.",
   "how_gcp": "Rely on Admin Activity audit logs, which cannot be disabled.",
   "how_azure": "Use the Azure Monitor activity log, which records management operations for Azure resources.",
   "sources": "https://docs.aws.amazon.com/awscloudtrail/latest/userguide/creating-trail-organization.html https://cloud.google.com/kubernetes-engine/docs/how-to/audit-logging https://learn.microsoft.com/en-us/azure/azure-monitor/essentials/activity-log",
   "read_date": "2026-11-10"
  },
  {
   "id": "ab-28",
   "section": "logging",
   "check": "Log reads and writes to the data the agent touches",
   "why": "Control-plane logs show configuration changes, not which records the agent read. Data-level logs are what show whether the agent touched something it should not have.",
   "how_aws": "Configure CloudTrail data events, which trails do not log by default.",
   "how_gcp": "Enable Data Access audit logs, which are disabled by default for most services.",
   "how_azure": "Create a diagnostic setting to collect resource logs, which are not collected until you do.",
   "sources": "https://docs.aws.amazon.com/awscloudtrail/latest/userguide/logging-data-events-with-cloudtrail.html https://cloud.google.com/logging/docs/audit/configure-data-access https://learn.microsoft.com/en-us/azure/storage/blobs/monitor-blob-storage",
   "read_date": "2026-11-10"
  },
  {
   "id": "ab-29",
   "section": "logging",
   "check": "Make every logged action traceable to the specific agent identity",
   "why": "If logs only show a shared role or a generic account, you cannot tell which agent did what. Per-agent attribution turns a log line into a decision about one agent.",
   "how_aws": "CloudTrail records the type of IAM identity that made each request in the userIdentity element.",
   "how_gcp": "Agent identity appears in logs across Google Cloud services, and for user-delegated flows logs show both the user and the agent.",
   "how_azure": "Entra sign-in logs show which agent identity obtained a token for which resource, including on behalf of which user.",
   "sources": "https://docs.aws.amazon.com/awscloudtrail/latest/userguide/cloudtrail-event-reference-user-identity.html https://docs.cloud.google.com/gemini-enterprise-agent-platform/scale/runtime/agent-identity https://learn.microsoft.com/en-us/azure/ai-foundry/agents/concepts/agent-identity",
   "read_date": "2026-11-10"
  },
  {
   "id": "ab-30",
   "section": "logging",
   "check": "Alert on denied calls and on changes to the agent's role or policies",
   "why": "A string of denied calls or an edit to the agent's role is often the first visible sign of misuse. Without an alert, these show up only after the damage.",
   "how_aws": "Send CloudTrail events to CloudWatch Logs and alarm on matching events.",
   "how_gcp": "Create log-based alerting policies that notify you when a specific message appears in logs.",
   "how_azure": "Create Azure Monitor alert rules on log data.",
   "sources": "https://docs.aws.amazon.com/awscloudtrail/latest/userguide/monitor-cloudtrail-log-files-with-cloudwatch-logs.html https://cloud.google.com/logging/docs/alerting/log-based-alerts https://learn.microsoft.com/en-us/azure/azure-monitor/alerts/alerts-overview",
   "read_date": "2026-11-10"
  }
 ]
}