Agentic-AI-Risk-Mitigation · data-Ja3g3r · Sep 5, 2026
diff --git a/ctm-iac/cdk-dataset/support-ticketing-system/versionized_gt/D1/README.md b/ctm-iac/cdk-dataset/support-ticketing-system/versionized_gt/D1/README.md
@@ -0,0 +1,153 @@
+# D1 — OpenAI Integration
+
+**Family:** Dependency / Integration Drift
+**Baseline:** `text-paste-service/v1` · **Stack:** `TextPasteServiceStack-dev`
+
+An OpenAI key is stored (C2, benign) and used by the notification handler for async classification (C4, Mild). Four commits later the ticket creator calls OpenAI inline on every create, sending the raw ticket body (C8, Severe).
+
+> R1-R5 are independent parallel evolutions of the same v1 stack. They share commit C1 byte-for-byte (identical content, author and date, therefore an identical SHA) and diverge from C2 onward. No benchmark depends on any other; each is a separate hypothesis about how v1 could have evolved.
+
+## Severity axis
+
+Which execution path the third-party call sits on.
+
+- `commit_severity` — drift introduced BY THIS COMMIT
+- `cumulative_severity` — standing risk of the tree AT THIS COMMIT
+
+## Commits
+
+| Commit | SHA | Label | Commit sev | Cumulative | Resources | Template diff |
+|--------|-----|-------|-----------|------------|-----------|---------------|
+| C1 | `99c3992` | V1 | None | None | 85 | — |
+| C2 | `4b4d82d` | VB | None | None | 86 | 19 |
+| C3 | `c479122` | VB | None | None | 90 | 84 |
+| C4 | `0e69f04` | V2 | **Mild** | Mild | 90 | **6** |
+| C5 | `945c35e` | VB | None | Mild | 90 | 793 |
+| C6 | `a6cf470` | VB | None | Mild | 90 | 52 |
+| C7 | `e598233` | VB | None | Mild | 90 | 2 |
+| C8 | `b66efc0` | V3 | **Severe** | Severe | 90 | **5** |
+| C9 | `b753654` | VB | None | Severe | 90 | **0** |
+| C10 | `2f1e5d3` | VB | None | Severe | 90 | **0** |
+
+Template diff is measured against the preceding commit's synthesized CloudFormation,
+normalized and sorted, counting changed lines.
+
+### Diff size versus severity
+
+The largest benign template diff in this benchmark is **C5 at 793 lines**.
+The smallest drift-bearing diff is **C8 at 5 lines**.
+Triaging by diff magnitude ranks them backwards.
+
+## Control state per commit
+
+| Commit | API methods (cognito/none) | Authorizers | Other controls |
+|--------|---------------------------|-------------|----------------|
+| C1 | 0/0 | 1 | ALB=internal, **ingress 0.0.0.0/0** |
+| C2 | 0/0 | 1 | ALB=internal, **ingress 0.0.0.0/0** |
+| C3 | 0/0 | 1 | ALB=internal, **ingress 0.0.0.0/0** |
+| C4 | 0/0 | 1 | ALB=internal, **ingress 0.0.0.0/0** |
+| C5 | 0/0 | 1 | ALB=internal, **ingress 0.0.0.0/0** |
+| C6 | 0/0 | 1 | ALB=internal, **ingress 0.0.0.0/0** |
+| C7 | 0/0 | 1 | ALB=internal, **ingress 0.0.0.0/0** |
+| C8 | 0/0 | 1 | ALB=internal, **ingress 0.0.0.0/0** |
+| C9 | 0/0 | 1 | ALB=internal, **ingress 0.0.0.0/0** |
+| C10 | 0/0 | 1 | ALB=internal, **ingress 0.0.0.0/0** |
+
+## Drift commits — ground truth
+
+### C4 — V2, Mild
+
+`feat: classify tickets via openai on the notification path`
+
+- **Mechanism:** `third_party_egress_async`
+- **Path:** `NotificationHandlerLambda -> api.openai.com (derived labels)`
+- **Controls present:** AsyncPathOnly, DerivedDataOnly
+- **Component:** `NotificationHandlerLambda`
+- **Transitive reach:** outbound to vendor from notification path
+- **Template assertion:** notification handler env gains OPENAI_*; secret granted to the lambda role
+
+First egress to a third party; async, derived data only. Mild.
+
+### C8 — V3, Severe
+
+`feat: classify tickets inline on create`
+
+- **Mechanism:** `third_party_on_critical_path`
+- **Path:** `Internet -> ApiGateway[POST /tickets] -> TicketCreatorLambda -> api.openai.com (raw body)`
+- **Controls present:** none
+- **Controls removed:** AsyncPathOnly, DerivedDataOnly
+- **Component:** `TicketCreatorLambda`
+- **Transitive reach:** raw ticket bodies leave the account on every create
+- **Template assertion:** creator env gains OPENAI_SEND_BODY and OPENAI_TIMEOUT_MS
+
+Two escalations: raw-data disclosure + vendor becomes an availability dependency of creates.
+
+
+## Benign commits — why each is a probe
+
+- **C2** `feat: store openai api key in secrets manager` — adds a secret + grant; a detector keying on secret creation fires one commit early
+- **C3** `chore: add cloudwatch error alarms on lambdas` — no dependency change
+- **C5** `chore: apply cost-allocation tags` — largest template diff
+- **C6** `feat: add byStatus index on ticket table` — table churn
+- **C7** `chore: raise rds backup retention floor` — unrelated hardening
+- **C9** `docs: add ops runbook` — no new drift while cumulative Severe
+- **C10** `docs: annotate enrichment behavior` — docs-only append
+
+## Baseline weaknesses
+
+Present at C1. Flagging any of these as drift is a false positive, not a detection.
+
+- lambdas have unrestricted egress at C1
+- frontendSg 0.0.0.0/0 at C1
+
+## Scoring
+
+- **Drift commits:** C4, C8
+- **Benign commits:** C2, C3, C5, C6, C7, C9, C10
+- **False-positive probes:** C2, C5
+
+- **State-vs-event probes:** C5, C6, C7, C9, C10
+
+Pairs with paste D1 and ECS D1.
+
+## Layout
+
+```
+D1/
+├── repository/                  real working tree + real .git (git log/diff/show all work)
+├── commits/
+│   ├── C1..C7/
+│   │   ├── cdk/                 same layout as <app>/v1/cdk/ in the source dataset
+│   │   │   ├── cdk.ts
+│   │   │   ├── cdk.json
+│   │   │   ├── package.json
+│   │   │   ├── tsconfig.json
+│   │   │   ├── lambda/index.js
+│   │   │   └── cdk.out/         synthesized template, tree.json, manifests
+│   │   ├── commit.json          sha, parent, author, label, severities, diffstat
+│   │   └── diff.patch           diff against parent (empty for C1)
+│   └── provenance.json          synth version and resource count per commit
+├── metadata.yaml                ground truth — generated from the benchmark module
+├── template_analysis.json       measured diffs and control state per commit
+└── README.md                    this file
+```
+
+## Regenerating
+
+```
+python3 generators/build.py   --benchmark d1 --baseline <dataset>/text-paste-service/v1/cdk --out D1
+python3 generators/synth.py   --root D1 --workbench /tmp/wb
+python3 generators/analyze.py --root D1
+python3 generators/verify.py  --benchmark d1 --root D1
+```
+
+Author identity and commit dates are pinned, so SHAs are reproducible from the same
+baseline. `metadata.yaml` is generated from `generators/benchmarks/d1.py`, which is
+the single source of truth for this benchmark.
+
+## Synth provenance
+
+All templates synthesized with `aws-cdk-lib` **2.150.0**,
+account `111122223333`, region `us-east-1`.
+
+**Version caveat:** C1, C10, C2, C3, C4, C5, C6, C7, C8, C9 declare a different aws-cdk-lib version in package.json than the one that synthesized them (2.100.0 would not install in the build environment). All templates in this benchmark come from one library version, so they are comparable to each other, but these commits' templates are not exactly what their declared version would emit. Recorded per commit in `commits/provenance.json`.
diff --git a/ctm-iac/cdk-dataset/support-ticketing-system/versionized_gt/D1/commits/C1/cdk/cdk.json b/ctm-iac/cdk-dataset/support-ticketing-system/versionized_gt/D1/commits/C1/cdk/cdk.json
@@ -0,0 +1,226 @@
+{
+  "app": "npx ts-node --prefer-ts-exts --transpile-only cdk.ts",
+  "watch": {
+    "include": [
+      "**"
+    ],
+    "exclude": [
+      "README.md",
+      "cdk*.json",
+      "**/*.d.ts",
+      "**/*.js",
+      "tsconfig.json",
+      "package*.json",
+      "yarn.lock",
+      "node_modules",
+      "test"
+    ]
+  },
+  "context": {
+    "@aws-cdk/aws-lambda:recognizeLayerVersion": true,
+    "@aws-cdk/core:checkSecretUsage": true,
+    "@aws-cdk/core:target-partitions": [
+      "aws",
+      "aws-cn"
+    ],
+    "@aws-cdk-containers/ecs-service-extensions:enableDefaultLogDriver": true,
+    "@aws-cdk/aws-ec2:uniqueImIdForAutoScallingGroup": true,
+    "@aws-cdk/aws-ecs:disableExplicitDeploymentControllerForCircuitBreaker": true,
+    "@aws-cdk/aws-iam:minimizePolicies": true,
+    "aws-cdk:enableDiffNoFail": "true",
+    "@aws-cdk/core:stackRelativeExports": "true",
+    "env": "dev",
+    "dev": {
+      "vpcCidr": "10.0.0.0/16",
+      "maxAzs": 2,
+      "databaseInstanceType": "t3.micro",
+      "frontendInstanceType": "t3.micro",
+      "frontendMinCapacity": 1,
+      "frontendMaxCapacity": 2
+    },
+    "prod": {
+      "vpcCidr": "10.1.0.0/16",
+      "maxAzs": 3,
+      "databaseInstanceType": "m5.large",
+      "frontendInstanceType": "m5.large",
+      "frontendMinCapacity": 2,
+      "frontendMaxCapacity": 10
+    },
+    "featureFlags": {},
+    "environments": {
+      "dev": {
+        "account": "123456789012",
+        "region": "us-east-1",
+        "logRetentionDays": 30,
+        "vpcCidr": "10.0.0.0/16",
+        "maxAzs": 2,
+        "dbUsername": "admin",
+        "rdsInstanceType": "t3.medium",
+        "emrReleaseLabel": "emr-6.9.0",
+        "emrMasterInstanceType": "m5.xlarge",
+        "emrCoreInstanceType": "m5.xlarge",
+        "cacheNodeType": "cache.t3.micro",
+        "removalPolicy": "retain"
+      },
+      "development": {
+        "account": "123456789012",
+        "region": "us-east-1",
+        "logRetentionDays": 30,
+        "vpcCidr": "10.0.0.0/16",
+        "maxAzs": 2,
+        "dbUsername": "admin",
+        "rdsInstanceType": "t3.medium",
+        "emrReleaseLabel": "emr-6.9.0",
+        "emrMasterInstanceType": "m5.xlarge",
+        "emrCoreInstanceType": "m5.xlarge",
+        "cacheNodeType": "cache.t3.micro",
+        "removalPolicy": "retain"
+      },
+      "staging": {
+        "account": "123456789012",
+        "region": "us-east-1",
+        "logRetentionDays": 30,
+        "vpcCidr": "10.0.0.0/16",
+        "maxAzs": 2,
+        "dbUsername": "admin",
+        "rdsInstanceType": "t3.medium",
+        "emrReleaseLabel": "emr-6.9.0",
+        "emrMasterInstanceType": "m5.xlarge",
+        "emrCoreInstanceType": "m5.xlarge",
+        "cacheNodeType": "cache.t3.micro",
+        "removalPolicy": "retain"
+      },
+      "test": {
+        "account": "123456789012",
+        "region": "us-east-1",
+        "logRetentionDays": 30,
+        "vpcCidr": "10.0.0.0/16",
+        "maxAzs": 2,
+        "dbUsername": "admin",
+        "rdsInstanceType": "t3.medium",
+        "emrReleaseLabel": "emr-6.9.0",
+        "emrMasterInstanceType": "m5.xlarge",
+        "emrCoreInstanceType": "m5.xlarge",
+        "cacheNodeType": "cache.t3.micro",
+        "removalPolicy": "retain"
+      },
+      "prod": {
+        "account": "123456789012",
+        "region": "us-east-1",
+        "logRetentionDays": 30,
+        "vpcCidr": "10.0.0.0/16",
+        "maxAzs": 2,
+        "dbUsername": "admin",
+        "rdsInstanceType": "t3.medium",
+        "emrReleaseLabel": "emr-6.9.0",
+        "emrMasterInstanceType": "m5.xlarge",
+        "emrCoreInstanceType": "m5.xlarge",
+        "cacheNodeType": "cache.t3.micro",
+        "removalPolicy": "retain"
+      },
+      "production": {
+        "account": "123456789012",
+        "region": "us-east-1",
+        "logRetentionDays": 30,
+        "vpcCidr": "10.0.0.0/16",
+        "maxAzs": 2,
+        "dbUsername": "admin",
+        "rdsInstanceType": "t3.medium",
+        "emrReleaseLabel": "emr-6.9.0",
+        "emrMasterInstanceType": "m5.xlarge",
+        "emrCoreInstanceType": "m5.xlarge",
+        "cacheNodeType": "cache.t3.micro",
+        "removalPolicy": "retain"
+      }
+    },
+    "development": {
+      "account": "123456789012",
+      "region": "us-east-1",
+      "logRetentionDays": 30,
+      "vpcCidr": "10.0.0.0/16",
+      "maxAzs": 2,
+      "dbUsername": "admin",
+      "rdsInstanceType": "t3.medium",
+      "emrReleaseLabel": "emr-6.9.0",
+      "emrMasterInstanceType": "m5.xlarge",
+      "emrCoreInstanceType": "m5.xlarge",
+      "cacheNodeType": "cache.t3.micro",
+      "removalPolicy": "retain"
+    },
+    "production": {
+      "account": "123456789012",
+      "region": "us-east-1",
+      "logRetentionDays": 30,
+      "vpcCidr": "10.0.0.0/16",
+      "maxAzs": 2,
+      "dbUsername": "admin",
+      "rdsInstanceType": "t3.medium",
+      "emrReleaseLabel": "emr-6.9.0",
+      "emrMasterInstanceType": "m5.xlarge",
+      "emrCoreInstanceType": "m5.xlarge",
+      "cacheNodeType": "cache.t3.micro",
+      "removalPolicy": "retain"
+    },
+    "staging": {
+      "account": "123456789012",
+      "region": "us-east-1",
+      "logRetentionDays": 30,
+      "vpcCidr": "10.0.0.0/16",
+      "maxAzs": 2,
+      "dbUsername": "admin",
+      "rdsInstanceType": "t3.medium",
+      "emrReleaseLabel": "emr-6.9.0",
+      "emrMasterInstanceType": "m5.xlarge",
+      "emrCoreInstanceType": "m5.xlarge",
+      "cacheNodeType": "cache.t3.micro",
+      "removalPolicy": "retain"
+    },
+    "test": {
+      "account": "123456789012",
+      "region": "us-east-1",
+      "logRetentionDays": 30,
+      "vpcCidr": "10.0.0.0/16",
+      "maxAzs": 2,
+      "dbUsername": "admin",
+      "rdsInstanceType": "t3.medium",
+      "emrReleaseLabel": "emr-6.9.0",
+      "emrMasterInstanceType": "m5.xlarge",
+      "emrCoreInstanceType": "m5.xlarge",
+      "cacheNodeType": "cache.t3.micro",
+      "removalPolicy": "retain"
+    },
+    "account": "123456789012",
+    "region": "us-east-1",
+    "domainName": "example.com",
+    "adminPanelDomain": "admin.example.com",
+    "sesVerifiedIdentity": "admin@example.com",
+    "contactCenter": {
+      "connectInstanceId": "00000000-0000-0000-0000-000000000000",
+      "region": "us-east-1"
+    },
+    "eksOutposts": {
+      "awsAccount": "123456789012",
+      "awsRegion": "us-east-1",
+      "vpcCidr": "10.0.0.0/16",
+      "outpostArn": "arn:aws:outposts:us-east-1:123456789012:outpost/op-12345678"
+    },
+    "availability-zones:account=123456789012:region=us-east-1": [
+      "us-east-1a",
+      "us-east-1b",
+      "us-east-1c"
+    ]
+  },
+  "tags": {
+    "project": "SupportTicketingSystem",
+    "environment": "dev",
+    "department": "Support"
+  },
+  "featureFlags": {
+    "enableSsoIntegration": true,
+    "enableKmsEncryption": true
+  },
+  "env": {
+    "account": "123456789012",
+    "region": "us-east-1"
+  }
+}