feat(aws-devops-agent): add the experiment-orchestration skill and agent prompt

README.md

@@ -60,6 +60,12 @@ Agent Skills are modular, text-based playbooks that teach an agent how to perfor
 | `metrics/launchdarkly-metric-create` | Create metrics and instrument tracking events |
 | `metrics/launchdarkly-metric-instrument` | Add tracking calls to code for existing metrics |
 
+### AWS DevOps Agent
+
+| Skill | Description |
+|-------|-------------|
+| `aws-devops-agent/experiment-orchestration` | Take a product goal through a LaunchDarkly experiment and a rollout (guarded, if the plan has it) in AWS DevOps Agent. `ldcli aws-devops-agent setup` installs it with the `experiment-orchestrator` custom agent. See [`skills/aws-devops-agent`](skills/aws-devops-agent/README.md). |
+
 ## Install as a Claude Code Plugin
 
 This repo is a [Claude Code plugin](https://code.claude.com/docs/en/create-plugins). Installing it gives you all the skills above plus the LaunchDarkly MCP server.

evals/README.md

@@ -34,6 +34,7 @@ npm run eval:agentcontrol-tools       # agentcontrol/tools
 npm run eval:configs-variations       # agentcontrol/configs-variations
 npm run eval:flag-create             # feature-flags/launchdarkly-flag-create
 npm run eval:onboarding              # onboarding
+npm run eval:experiment-orchestration # aws-devops-agent/experiment-orchestration
 
 # Quick smoke check — first test case only (~15-20s, ~$0.05)
 npm run eval:configs-create:single
@@ -42,6 +43,7 @@ npm run eval:agentcontrol-tools:single
 npm run eval:configs-variations:single
 npm run eval:flag-create:single
 npm run eval:onboarding:single
+npm run eval:experiment-orchestration:single
 
 # Aggregate and CI operations
 npm run eval:all                     # Run every suite and rebuild ../eval-scores.json

evals/experiment-orchestration/mocks.json

@@ -0,0 +1,113 @@
+{
+  "start-guarded-rollout": {
+    "flagKey": "{{flagKey}}",
+    "environmentKey": "{{environmentKey}}",
+    "status": "monitoring",
+    "currentStage": 1,
+    "stageCount": 3
+  },
+  "update-rollout": {
+    "key": "{{flagKey}}",
+    "environment": "{{environmentKey}}",
+    "fallthrough": {
+      "variation": 0
+    }
+  },
+  "stop-experiment-iteration": {
+    "key": "{{experimentKey}}",
+    "environmentKey": "{{environmentKey}}",
+    "currentIteration": {
+      "status": "stopped"
+    }
+  },
+  "get-experiment-results": {
+    "experimentKey": "{{experimentKey}}",
+    "metrics": [
+      {
+        "metricKey": "checkout-completed",
+        "isPrimary": true,
+        "treatments": [
+          {
+            "treatmentId": "treat-control",
+            "name": "Control",
+            "isBest": false,
+            "analyzedUnitCount": 4200
+          },
+          {
+            "treatmentId": "treat-treatment",
+            "name": "Treatment",
+            "isBest": true,
+            "relativeDifference": 0.081,
+            "isSignificant": true,
+            "analyzedUnitCount": 4180
+          }
+        ]
+      }
+    ]
+  },
+  "get-experiment-metric-results": {
+    "experimentKey": "{{experimentKey}}",
+    "metricKey": "{{metricKey}}",
+    "treatments": [
+      {
+        "treatmentId": "treat-control",
+        "name": "Control",
+        "analyzedUnitCount": 4200,
+        "trafficCount": 4200,
+        "conversionCount": 630,
+        "mean": 0.15
+      },
+      {
+        "treatmentId": "treat-treatment",
+        "name": "Treatment",
+        "analyzedUnitCount": 4180,
+        "trafficCount": 4180,
+        "conversionCount": 678,
+        "mean": 0.162,
+        "statistics": {
+          "relativeDifferences": [
+            {
+              "estimate": 0.081,
+              "lower": 0.02,
+              "upper": 0.14,
+              "pValue": 0.008
+            }
+          ],
+          "isSignificant": true
+        }
+      }
+    ]
+  },
+  "create-approval-request": {
+    "_id": "apr-0001",
+    "flagKey": "{{flagKey}}",
+    "environmentKey": "{{environmentKey}}",
+    "status": "pending",
+    "reviewStatus": "pending"
+  },
+  "apply-approval-request": {
+    "_id": "apr-0001",
+    "status": "completed",
+    "reviewStatus": "approved"
+  },
+  "list-approval-requests": {
+    "items": [
+      {
+        "_id": "apr-0001",
+        "flagKey": "{{flagKey}}",
+        "status": "pending",
+        "reviewStatus": "pending"
+      }
+    ],
+    "totalCount": 1
+  },
+  "list-metric-events": {
+    "events": [
+      {
+        "eventKey": "checkout-completed",
+        "lastSeen": "2026-10-28T12:00:00Z"
+      }
+    ],
+    "totalCount": 1
+  }
+}