@adcp/client 4.23.0 → 4.24.0

package/storyboards/deterministic_testing.yaml

@@ -1,8 +1,8 @@
 id: deterministic_testing
-version: "1.0.0"
-title: "Deterministic testing"
+version: '1.0.0'
+title: 'Deterministic testing'
 category: core
-summary: "Uses comply_test_controller to force state transitions and simulate delivery/budget, verifying state machines and reporting."
+summary: 'Uses comply_test_controller to force state transitions and simulate delivery/budget, verifying state machines and reporting.'
 track: core
 required_tools:
   - comply_test_controller
@@ -36,22 +36,22 @@ agent:
     - sells_media
     - supports_test_controller
   examples:
-    - "Any seller with comply_test_controller"
+    - 'Any seller with comply_test_controller'
 
 caller:
   role: buyer_agent
-  example: "Comply test harness"
+  example: 'Comply test harness'
 
 prerequisites:
   description: |
     The seller must expose the comply_test_controller tool. The controller is detected
     via list_scenarios before any state transitions are attempted. Phases that require
     specific scenarios are skipped if the controller does not support them.
-  test_kit: "test-kits/acme-outdoor.yaml"
+  test_kit: 'test-kits/acme-outdoor.yaml'
 
 phases:
   - id: controller_validation
-    title: "Controller validation"
+    title: 'Controller validation'
     narrative: |
       Validates that the comply_test_controller is functional and responds correctly
       to scenario discovery, unknown scenarios, missing parameters, and nonexistent
@@ -59,7 +59,7 @@ phases:
 
     steps:
       - id: list_scenarios
-        title: "List supported scenarios"
+        title: 'List supported scenarios'
         requires_tool: comply_test_controller
         narrative: |
           Call the controller with scenario: list_scenarios to discover which
@@ -69,27 +69,27 @@ phases:
         stateful: false
         context_outputs:
           - name: supported_scenarios
-            path: "scenarios"
+            path: 'scenarios'
         expected: |
           Return a list of supported scenarios. The response should include:
           - success: true
           - scenarios: array or object of scenario names
 
         sample_request:
-          scenario: "list_scenarios"
+          scenario: 'list_scenarios'
 
         validations:
           - check: field_present
-            path: "scenarios"
-            description: "Controller returns supported scenarios"
+            path: 'scenarios'
+            description: 'Controller returns supported scenarios'
           - check: field_value
-            path: "success"
+            path: 'success'
             allowed_values:
               - true
-            description: "list_scenarios succeeds"
+            description: 'list_scenarios succeeds'
 
       - id: unknown_scenario
-        title: "Unknown scenario returns error"
+        title: 'Unknown scenario returns error'
         requires_tool: comply_test_controller
         narrative: |
           Send a nonexistent scenario name. The controller should return
@@ -97,29 +97,30 @@ phases:
         task: comply_test_controller
         comply_scenario: controller_validation
         stateful: false
+        expect_error: true
         expected: |
           Return an error response with:
           - success: false
           - error: UNKNOWN_SCENARIO
 
         sample_request:
-          scenario: "nonexistent_scenario"
+          scenario: 'nonexistent_scenario'
           params: {}
 
         validations:
           - check: field_value
-            path: "success"
+            path: 'success'
             allowed_values:
               - false
-            description: "Unknown scenario fails"
+            description: 'Unknown scenario fails'
           - check: field_value
-            path: "error"
+            path: 'error'
             allowed_values:
-              - "UNKNOWN_SCENARIO"
-            description: "Error code is UNKNOWN_SCENARIO"
+              - 'UNKNOWN_SCENARIO'
+            description: 'Error code is UNKNOWN_SCENARIO'
 
       - id: missing_params
-        title: "Missing params returns error"
+        title: 'Missing params returns error'
         requires_tool: comply_test_controller
         narrative: |
           Call force_creative_status with empty params. The controller should
@@ -127,29 +128,30 @@ phases:
         task: comply_test_controller
         comply_scenario: controller_validation
         stateful: false
+        expect_error: true
         expected: |
           Return an error response with:
           - success: false
           - error: INVALID_PARAMS
 
         sample_request:
-          scenario: "force_creative_status"
+          scenario: 'force_creative_status'
           params: {}
 
         validations:
           - check: field_value
-            path: "success"
+            path: 'success'
             allowed_values:
               - false
-            description: "Missing params fails"
+            description: 'Missing params fails'
           - check: field_value
-            path: "error"
+            path: 'error'
             allowed_values:
-              - "INVALID_PARAMS"
-            description: "Error code is INVALID_PARAMS"
+              - 'INVALID_PARAMS'
+            description: 'Error code is INVALID_PARAMS'
 
       - id: not_found_entity
-        title: "Nonexistent entity returns NOT_FOUND"
+        title: 'Nonexistent entity returns NOT_FOUND'
         requires_tool: comply_test_controller
         narrative: |
           Call force_creative_status with a nonexistent creative_id. The controller
@@ -157,39 +159,61 @@ phases:
         task: comply_test_controller
         comply_scenario: controller_validation
         stateful: false
+        expect_error: true
         expected: |
           Return an error response with:
           - success: false
           - error: NOT_FOUND
 
         sample_request:
-          scenario: "force_creative_status"
+          scenario: 'force_creative_status'
           params:
-            creative_id: "comply-test-nonexistent-000000000000"
-            status: "approved"
+            creative_id: 'comply-test-nonexistent-000000000000'
+            status: 'approved'
 
         validations:
           - check: field_value
-            path: "success"
+            path: 'success'
             allowed_values:
               - false
-            description: "Nonexistent entity fails"
+            description: 'Nonexistent entity fails'
           - check: field_value
-            path: "error"
+            path: 'error'
             allowed_values:
-              - "NOT_FOUND"
-            description: "Error code is NOT_FOUND"
+              - 'NOT_FOUND'
+            description: 'Error code is NOT_FOUND'
 
   - id: deterministic_account
-    title: "Deterministic account state machine"
+    title: 'Deterministic account state machine'
     narrative: |
       Forces account status transitions via the controller and verifies them through
       list_accounts. Tests suspension gating (operations blocked when suspended),
       payment_required state, and reactivation.
 
     steps:
+      - id: sync_accounts_for_state
+        title: 'Create sandbox account for state machine test'
+        requires_tool: comply_test_controller
+        narrative: |
+          Sync a sandbox account so the state machine has an entity to
+          force transitions on.
+        task: sync_accounts
+        comply_scenario: deterministic_account
+        stateful: true
+
+        sample_request:
+          brand:
+            domain: 'test.example.com'
+          operator: 'test.example.com'
+          sandbox: true
+
+        validations:
+          - check: field_present
+            path: 'accounts[0].account_id'
+            description: 'Account sync returns account_id'
+
       - id: list_accounts_for_state
-        title: "Find account for state machine test"
+        title: 'Find account for state machine test'
         requires_tool: comply_test_controller
         narrative: |
           List accounts to find an active account for state transitions.
@@ -198,7 +222,7 @@ phases:
         stateful: false
         context_outputs:
           - name: account_id
-            path: "accounts[0].account_id"
+            path: 'accounts[0].account_id'
         expected: |
           Return at least one account with an account_id.
 
@@ -206,11 +230,11 @@ phases:
 
         validations:
           - check: field_present
-            path: "accounts[0].account_id"
-            description: "At least one account exists"
+            path: 'accounts[0].account_id'
+            description: 'At least one account exists'
 
       - id: force_account_suspended
-        title: "Force account to suspended"
+        title: 'Force account to suspended'
         requires_tool: comply_test_controller
         narrative: |
           Use the controller to force the account into suspended state. Operations
@@ -225,25 +249,25 @@ phases:
           - current_state: suspended
 
         sample_request:
-          scenario: "force_account_status"
+          scenario: 'force_account_status'
           params:
-            account_id: "{{account_id}}"
-            status: "suspended"
+            account_id: '$context.account_id'
+            status: 'suspended'
 
         validations:
           - check: field_value
-            path: "success"
+            path: 'success'
             allowed_values:
               - true
-            description: "Account suspension succeeds"
+            description: 'Account suspension succeeds'
           - check: field_value
-            path: "current_state"
+            path: 'current_state'
             allowed_values:
-              - "suspended"
-            description: "Account is now suspended"
+              - 'suspended'
+            description: 'Account is now suspended'
 
       - id: force_account_active
-        title: "Reactivate account"
+        title: 'Reactivate account'
         requires_tool: comply_test_controller
         narrative: |
           Restore the account to active state. Verify the transition from suspended
@@ -258,25 +282,25 @@ phases:
           - current_state: active
 
         sample_request:
-          scenario: "force_account_status"
+          scenario: 'force_account_status'
           params:
-            account_id: "{{account_id}}"
-            status: "active"
+            account_id: '$context.account_id'
+            status: 'active'
 
         validations:
           - check: field_value
-            path: "success"
+            path: 'success'
             allowed_values:
               - true
-            description: "Account reactivation succeeds"
+            description: 'Account reactivation succeeds'
           - check: field_value
-            path: "current_state"
+            path: 'current_state'
             allowed_values:
-              - "active"
-            description: "Account is now active"
+              - 'active'
+            description: 'Account is now active'
 
       - id: force_account_payment_required
-        title: "Force account to payment_required"
+        title: 'Force account to payment_required'
         requires_tool: comply_test_controller
         narrative: |
           Force the account into payment_required state. This tests a distinct
@@ -288,20 +312,20 @@ phases:
           Return a successful state transition to payment_required.
 
         sample_request:
-          scenario: "force_account_status"
+          scenario: 'force_account_status'
           params:
-            account_id: "{{account_id}}"
-            status: "payment_required"
+            account_id: '$context.account_id'
+            status: 'payment_required'
 
         validations:
           - check: field_value
-            path: "success"
+            path: 'success'
             allowed_values:
               - true
-            description: "Transition to payment_required succeeds"
+            description: 'Transition to payment_required succeeds'
 
       - id: restore_account_active
-        title: "Restore account to active"
+        title: 'Restore account to active'
         requires_tool: comply_test_controller
         narrative: |
           Restore the account to active state after the payment_required test.
@@ -312,20 +336,20 @@ phases:
           Account restored to active state.
 
         sample_request:
-          scenario: "force_account_status"
+          scenario: 'force_account_status'
           params:
-            account_id: "{{account_id}}"
-            status: "active"
+            account_id: '$context.account_id'
+            status: 'active'
 
         validations:
           - check: field_value
-            path: "success"
+            path: 'success'
             allowed_values:
               - true
-            description: "Account restored to active"
+            description: 'Account restored to active'
 
   - id: deterministic_media_buy
-    title: "Deterministic media buy state machine"
+    title: 'Deterministic media buy state machine'
     narrative: |
       Creates a media buy through the normal flow, then uses the controller to force
       it through status transitions: active, completed (terminal). Verifies that
@@ -333,43 +357,43 @@ phases:
 
     steps:
       - id: create_media_buy
-        title: "Create media buy for state machine test"
+        title: 'Create media buy for state machine test'
         requires_tool: comply_test_controller
         narrative: |
           Create a media buy through the standard create_media_buy task. The resulting
           media_buy_id is used for controller-driven state transitions.
         task: create_media_buy
-        schema_ref: "media-buy/create-media-buy-request.json"
-        response_schema_ref: "media-buy/create-media-buy-response.json"
-        doc_ref: "/media-buy/task-reference/create_media_buy"
+        schema_ref: 'media-buy/create-media-buy-request.json'
+        response_schema_ref: 'media-buy/create-media-buy-response.json'
+        doc_ref: '/media-buy/task-reference/create_media_buy'
         comply_scenario: deterministic_media_buy
         stateful: true
         context_outputs:
           - name: media_buy_id
-            path: "media_buy_id"
+            path: 'media_buy_id'
         expected: |
           Create a media buy and return a media_buy_id for state machine testing.
 
         sample_request:
           account:
             brand:
-              domain: "acmeoutdoor.com"
-            operator: "pinnacle-agency.com"
+              domain: 'acmeoutdoor.com'
+            operator: 'pinnacle-agency.com'
           brand:
-            domain: "acmeoutdoor.com"
-          start_time: "2026-04-01T00:00:00Z"
-          end_time: "2026-06-30T23:59:59Z"
+            domain: 'acmeoutdoor.com'
+          start_time: '2026-04-01T00:00:00Z'
+          end_time: '2026-06-30T23:59:59Z'
           packages:
-            - product_id: "test-product"
+            - product_id: 'test-product'
               budget: 5000
-              pricing_option_id: "test-pricing"
+              pricing_option_id: 'test-pricing'
 
         validations:
           - check: response_schema
-            description: "Response matches create-media-buy-response.json schema"
+            description: 'Response matches create-media-buy-response.json schema'
 
       - id: force_media_buy_active
-        title: "Force media buy to active"
+        title: 'Force media buy to active'
         requires_tool: comply_test_controller
         narrative: |
           Use the controller to force the media buy into active state.
@@ -380,33 +404,33 @@ phases:
           Return a successful state transition to active.
 
         sample_request:
-          scenario: "force_media_buy_status"
+          scenario: 'force_media_buy_status'
           params:
-            media_buy_id: "{{media_buy_id}}"
-            status: "active"
+            media_buy_id: '$context.media_buy_id'
+            status: 'active'
 
         validations:
           - check: field_value
-            path: "success"
+            path: 'success'
             allowed_values:
               - true
-            description: "Media buy activation succeeds"
+            description: 'Media buy activation succeeds'
           - check: field_value
-            path: "current_state"
+            path: 'current_state'
             allowed_values:
-              - "active"
-            description: "Media buy is now active"
+              - 'active'
+            description: 'Media buy is now active'
 
       - id: verify_media_buy_active
-        title: "Verify media buy status via get_media_buys"
+        title: 'Verify media buy status via get_media_buys'
         requires_tool: comply_test_controller
         narrative: |
           Call get_media_buys to confirm the controller-forced status is reflected
           in the standard API.
         task: get_media_buys
-        schema_ref: "media-buy/get-media-buys-request.json"
-        response_schema_ref: "media-buy/get-media-buys-response.json"
-        doc_ref: "/media-buy/task-reference/get_media_buys"
+        schema_ref: 'media-buy/get-media-buys-request.json'
+        response_schema_ref: 'media-buy/get-media-buys-response.json'
+        doc_ref: '/media-buy/task-reference/get_media_buys'
         comply_scenario: deterministic_media_buy
         stateful: true
         expected: |
@@ -415,20 +439,20 @@ phases:
         sample_request:
           account:
             brand:
-              domain: "acmeoutdoor.com"
-            operator: "pinnacle-agency.com"
+              domain: 'acmeoutdoor.com'
+            operator: 'pinnacle-agency.com'
           media_buy_ids:
-            - "{{media_buy_id}}"
+            - '$context.media_buy_id'
 
         validations:
           - check: response_schema
-            description: "Response matches get-media-buys-response.json schema"
+            description: 'Response matches get-media-buys-response.json schema'
           - check: field_present
-            path: "media_buys[0].status"
-            description: "Media buy has a status field"
+            path: 'media_buys[0].status'
+            description: 'Media buy has a status field'
 
       - id: force_media_buy_completed
-        title: "Force media buy to completed (terminal)"
+        title: 'Force media buy to completed (terminal)'
         requires_tool: comply_test_controller
         narrative: |
           Force the media buy to completed, a terminal state. No further transitions
@@ -440,25 +464,25 @@ phases:
           Return a successful transition to completed.
 
         sample_request:
-          scenario: "force_media_buy_status"
+          scenario: 'force_media_buy_status'
           params:
-            media_buy_id: "{{media_buy_id}}"
-            status: "completed"
+            media_buy_id: '$context.media_buy_id'
+            status: 'completed'
 
         validations:
           - check: field_value
-            path: "success"
+            path: 'success'
             allowed_values:
               - true
-            description: "Transition to completed succeeds"
+            description: 'Transition to completed succeeds'
           - check: field_value
-            path: "current_state"
+            path: 'current_state'
             allowed_values:
-              - "completed"
-            description: "Media buy is now completed"
+              - 'completed'
+            description: 'Media buy is now completed'
 
       - id: invalid_transition_from_terminal
-        title: "Reject transition from terminal state"
+        title: 'Reject transition from terminal state'
         requires_tool: comply_test_controller
         narrative: |
           Attempt to force the completed media buy back to active. The controller
@@ -466,31 +490,32 @@ phases:
         task: comply_test_controller
         comply_scenario: deterministic_media_buy
         stateful: true
+        expect_error: true
         expected: |
           Return an error response:
           - success: false
           - error: INVALID_TRANSITION
 
         sample_request:
-          scenario: "force_media_buy_status"
+          scenario: 'force_media_buy_status'
           params:
-            media_buy_id: "{{media_buy_id}}"
-            status: "active"
+            media_buy_id: '$context.media_buy_id'
+            status: 'active'
 
         validations:
           - check: field_value
-            path: "success"
+            path: 'success'
             allowed_values:
               - false
-            description: "Terminal state rejects transition"
+            description: 'Terminal state rejects transition'
           - check: field_value
-            path: "error"
+            path: 'error'
             allowed_values:
-              - "INVALID_TRANSITION"
-            description: "Error code is INVALID_TRANSITION"
+              - 'INVALID_TRANSITION'
+            description: 'Error code is INVALID_TRANSITION'
 
   - id: deterministic_creative
-    title: "Deterministic creative state machine"
+    title: 'Deterministic creative state machine'
     narrative: |
       Syncs a creative, then uses the controller to force it through status
       transitions: approved, archived (terminal). Verifies terminal state rejection
@@ -498,46 +523,46 @@ phases:
 
     steps:
       - id: sync_creative_for_state
-        title: "Sync creative for state machine test"
+        title: 'Sync creative for state machine test'
         requires_tool: comply_test_controller
         narrative: |
           Push a test creative via sync_creatives. The creative_id is captured
           for controller-driven transitions.
         task: sync_creatives
-        schema_ref: "creative/sync-creatives-request.json"
-        response_schema_ref: "creative/sync-creatives-response.json"
-        doc_ref: "/creative/task-reference/sync_creatives"
+        schema_ref: 'creative/sync-creatives-request.json'
+        response_schema_ref: 'creative/sync-creatives-response.json'
+        doc_ref: '/creative/task-reference/sync_creatives'
         comply_scenario: deterministic_creative
         stateful: true
         context_outputs:
           - name: creative_id
-            path: "results[0].creative_id"
+            path: 'creatives[0].creative_id'
         expected: |
           Accept the creative and return a creative_id for state machine testing.
 
         sample_request:
           account:
             brand:
-              domain: "acmeoutdoor.com"
-            operator: "pinnacle-agency.com"
+              domain: 'acmeoutdoor.com'
+            operator: 'pinnacle-agency.com'
           creatives:
-            - creative_id: "comply-state-test-creative"
-              name: "Comply State Test Creative"
+            - creative_id: 'comply-state-test-creative'
+              name: 'Comply State Test Creative'
               format_id:
-                agent_url: "https://your-platform.example.com"
-                id: "display_300x250"
+                agent_url: 'https://your-platform.example.com'
+                id: 'display_300x250'
               assets:
-                - asset_id: "image"
-                  asset_type: "image"
-                  url: "https://via.placeholder.com/300x250"
-                  mime_type: "image/png"
+                - asset_id: 'image'
+                  asset_type: 'image'
+                  url: 'https://via.placeholder.com/300x250'
+                  mime_type: 'image/png'
 
         validations:
           - check: response_schema
-            description: "Response matches sync-creatives-response.json schema"
+            description: 'Response matches sync-creatives-response.json schema'
 
       - id: force_creative_approved
-        title: "Force creative to approved"
+        title: 'Force creative to approved'
         requires_tool: comply_test_controller
         narrative: |
           Use the controller to force the creative into approved state.
@@ -548,25 +573,25 @@ phases:
           Successful transition to approved.
 
         sample_request:
-          scenario: "force_creative_status"
+          scenario: 'force_creative_status'
           params:
-            creative_id: "{{creative_id}}"
-            status: "approved"
+            creative_id: '$context.creative_id'
+            status: 'approved'
 
         validations:
           - check: field_value
-            path: "success"
+            path: 'success'
             allowed_values:
               - true
-            description: "Creative approval succeeds"
+            description: 'Creative approval succeeds'
           - check: field_value
-            path: "current_state"
+            path: 'current_state'
             allowed_values:
-              - "approved"
-            description: "Creative is now approved"
+              - 'approved'
+            description: 'Creative is now approved'
 
       - id: force_creative_archived
-        title: "Force creative to archived (terminal)"
+        title: 'Force creative to archived (terminal)'
         requires_tool: comply_test_controller
         narrative: |
           Force the creative to archived, a terminal state.
@@ -577,20 +602,20 @@ phases:
           Successful transition to archived.
 
         sample_request:
-          scenario: "force_creative_status"
+          scenario: 'force_creative_status'
           params:
-            creative_id: "{{creative_id}}"
-            status: "archived"
+            creative_id: '$context.creative_id'
+            status: 'archived'
 
         validations:
           - check: field_value
-            path: "success"
+            path: 'success'
             allowed_values:
               - true
-            description: "Transition to archived succeeds"
+            description: 'Transition to archived succeeds'
 
       - id: invalid_creative_transition
-        title: "Reject archived to processing"
+        title: 'Reject archived to processing'
         requires_tool: comply_test_controller
         narrative: |
           Attempt to force the archived creative back to processing. The controller
@@ -598,29 +623,30 @@ phases:
         task: comply_test_controller
         comply_scenario: deterministic_creative
         stateful: true
+        expect_error: true
         expected: |
           Return error with INVALID_TRANSITION.
 
         sample_request:
-          scenario: "force_creative_status"
+          scenario: 'force_creative_status'
           params:
-            creative_id: "{{creative_id}}"
-            status: "processing"
+            creative_id: '$context.creative_id'
+            status: 'processing'
 
         validations:
           - check: field_value
-            path: "success"
+            path: 'success'
             allowed_values:
               - false
-            description: "Terminal state rejects transition"
+            description: 'Terminal state rejects transition'
           - check: field_value
-            path: "error"
+            path: 'error'
             allowed_values:
-              - "INVALID_TRANSITION"
-            description: "Error code is INVALID_TRANSITION"
+              - 'INVALID_TRANSITION'
+            description: 'Error code is INVALID_TRANSITION'
 
       - id: force_creative_rejected
-        title: "Force fresh creative to rejected with reason"
+        title: 'Force fresh creative to rejected with reason'
         requires_tool: comply_test_controller
         narrative: |
           Sync a fresh creative, force it to pending_review, then reject it with a
@@ -632,29 +658,29 @@ phases:
           Successful transition to rejected with reason preserved.
 
         sample_request:
-          scenario: "force_creative_status"
+          scenario: 'force_creative_status'
           params:
-            creative_id: "{{creative_id}}"
-            status: "rejected"
-            rejection_reason: "Brand safety policy violation (comply test)"
+            creative_id: '$context.creative_id'
+            status: 'rejected'
+            rejection_reason: 'Brand safety policy violation (comply test)'
 
         validations:
           - check: field_value
-            path: "success"
+            path: 'success'
             allowed_values:
               - true
-            description: "Creative rejection succeeds"
+            description: 'Creative rejection succeeds'
 
   - id: deterministic_session
-    title: "Deterministic SI session state machine"
+    title: 'Deterministic SI session state machine'
     narrative: |
       Initiates an SI session, then uses the controller to force session timeout.
       Verifies that messaging on a terminated session fails appropriately.
 
     steps:
       - id: initiate_session
-        title: "Initiate SI session"
-        requires_tool: comply_test_controller
+        title: 'Initiate SI session'
+        requires_tool: si_initiate_session
         narrative: |
           Create an SI session via si_initiate_session. The session_id is captured
           for controller-driven termination.
@@ -663,24 +689,24 @@ phases:
         stateful: true
         context_outputs:
           - name: session_id
-            path: "session_id"
+            path: 'session_id'
         expected: |
           Return a session with a session_id.
 
         sample_request:
           identity:
-            user_type: "consumer"
+            user_type: 'consumer'
           supported_capabilities:
             response_formats:
-              - "text"
+              - 'text'
 
         validations:
           - check: field_present
-            path: "session_id"
-            description: "Session has a session_id"
+            path: 'session_id'
+            description: 'Session has a session_id'
 
       - id: force_session_terminated
-        title: "Force session timeout"
+        title: 'Force session timeout'
         requires_tool: comply_test_controller
         narrative: |
           Use the controller to force the session into terminated state with
@@ -692,21 +718,21 @@ phases:
           Successful transition to terminated.
 
         sample_request:
-          scenario: "force_session_status"
+          scenario: 'force_session_status'
           params:
-            session_id: "{{session_id}}"
-            status: "terminated"
-            termination_reason: "session_timeout"
+            session_id: '$context.session_id'
+            status: 'terminated'
+            termination_reason: 'session_timeout'
 
         validations:
           - check: field_value
-            path: "success"
+            path: 'success'
             allowed_values:
               - true
-            description: "Session termination succeeds"
+            description: 'Session termination succeeds'
 
       - id: verify_terminated_session
-        title: "Verify messaging fails on terminated session"
+        title: 'Verify messaging fails on terminated session'
         requires_tool: comply_test_controller
         narrative: |
           Attempt to send a message on the terminated session. The agent should
@@ -719,60 +745,60 @@ phases:
           must not accept messages on a terminated session.
 
         sample_request:
-          session_id: "{{session_id}}"
-          message: "comply test — session should be terminated"
+          session_id: '$context.session_id'
+          message: 'comply test — session should be terminated'
 
         validations:
           - check: field_value
-            path: "success"
+            path: 'success'
             allowed_values:
               - false
-            description: "Message on terminated session fails"
+            description: 'Message on terminated session fails'
 
   - id: deterministic_delivery
-    title: "Deterministic delivery simulation"
+    title: 'Deterministic delivery simulation'
     narrative: |
       Creates a media buy, simulates delivery data via the controller, then verifies
       the simulated metrics are reflected in get_media_buy_delivery.
 
     steps:
       - id: create_media_buy_for_delivery
-        title: "Create media buy for delivery test"
+        title: 'Create media buy for delivery test'
         requires_tool: comply_test_controller
         narrative: |
           Create a media buy to receive simulated delivery data.
         task: create_media_buy
-        schema_ref: "media-buy/create-media-buy-request.json"
-        response_schema_ref: "media-buy/create-media-buy-response.json"
-        doc_ref: "/media-buy/task-reference/create_media_buy"
+        schema_ref: 'media-buy/create-media-buy-request.json'
+        response_schema_ref: 'media-buy/create-media-buy-response.json'
+        doc_ref: '/media-buy/task-reference/create_media_buy'
         comply_scenario: deterministic_delivery
         stateful: true
         context_outputs:
           - name: delivery_media_buy_id
-            path: "media_buy_id"
+            path: 'media_buy_id'
         expected: |
           Create a media buy and return a media_buy_id.
 
         sample_request:
           account:
             brand:
-              domain: "acmeoutdoor.com"
-            operator: "pinnacle-agency.com"
+              domain: 'acmeoutdoor.com'
+            operator: 'pinnacle-agency.com'
           brand:
-            domain: "acmeoutdoor.com"
-          start_time: "2026-04-01T00:00:00Z"
-          end_time: "2026-06-30T23:59:59Z"
+            domain: 'acmeoutdoor.com'
+          start_time: '2026-04-01T00:00:00Z'
+          end_time: '2026-06-30T23:59:59Z'
           packages:
-            - product_id: "test-product"
+            - product_id: 'test-product'
               budget: 5000
-              pricing_option_id: "test-pricing"
+              pricing_option_id: 'test-pricing'
 
         validations:
           - check: response_schema
-            description: "Response matches create-media-buy-response.json schema"
+            description: 'Response matches create-media-buy-response.json schema'
 
       - id: simulate_delivery
-        title: "Simulate delivery data"
+        title: 'Simulate delivery data'
         requires_tool: comply_test_controller
         narrative: |
           Inject simulated delivery metrics (impressions, clicks, spend) into the
@@ -784,32 +810,32 @@ phases:
           Return success with simulated delivery metrics acknowledged.
 
         sample_request:
-          scenario: "simulate_delivery"
+          scenario: 'simulate_delivery'
           params:
-            media_buy_id: "{{delivery_media_buy_id}}"
+            media_buy_id: '$context.delivery_media_buy_id'
             impressions: 10000
             clicks: 150
             reported_spend:
               amount: 150.00
-              currency: "USD"
+              currency: 'USD'
 
         validations:
           - check: field_value
-            path: "success"
+            path: 'success'
             allowed_values:
               - true
-            description: "Delivery simulation succeeds"
+            description: 'Delivery simulation succeeds'
 
       - id: verify_delivery
-        title: "Verify delivery via get_media_buy_delivery"
+        title: 'Verify delivery via get_media_buy_delivery'
         requires_tool: comply_test_controller
         narrative: |
           Call get_media_buy_delivery and verify the simulated metrics are reflected
           in the response. Expect at least 10000 impressions.
         task: get_media_buy_delivery
-        schema_ref: "media-buy/get-media-buy-delivery-request.json"
-        response_schema_ref: "media-buy/get-media-buy-delivery-response.json"
-        doc_ref: "/media-buy/task-reference/get_media_buy_delivery"
+        schema_ref: 'media-buy/get-media-buy-delivery-request.json'
+        response_schema_ref: 'media-buy/get-media-buy-delivery-response.json'
+        doc_ref: '/media-buy/task-reference/get_media_buy_delivery'
         comply_scenario: deterministic_delivery
         stateful: true
         expected: |
@@ -820,62 +846,62 @@ phases:
         sample_request:
           account:
             brand:
-              domain: "acmeoutdoor.com"
-            operator: "pinnacle-agency.com"
+              domain: 'acmeoutdoor.com'
+            operator: 'pinnacle-agency.com'
           media_buy_ids:
-            - "{{delivery_media_buy_id}}"
+            - '$context.delivery_media_buy_id'
 
         validations:
           - check: response_schema
-            description: "Response matches get-media-buy-delivery-response.json schema"
+            description: 'Response matches get-media-buy-delivery-response.json schema'
           - check: field_present
-            path: "media_buys"
-            description: "Response contains delivery data"
+            path: 'media_buy_deliveries'
+            description: 'Response contains delivery data'
 
   - id: deterministic_budget
-    title: "Deterministic budget simulation"
+    title: 'Deterministic budget simulation'
     narrative: |
       Creates a media buy, then uses the controller to simulate budget spend at 95%
       and 100%. Verifies the agent handles near-depletion and full depletion.
 
     steps:
       - id: create_media_buy_for_budget
-        title: "Create media buy for budget test"
+        title: 'Create media buy for budget test'
         requires_tool: comply_test_controller
         narrative: |
           Create a media buy with a known budget for spend simulation.
         task: create_media_buy
-        schema_ref: "media-buy/create-media-buy-request.json"
-        response_schema_ref: "media-buy/create-media-buy-response.json"
-        doc_ref: "/media-buy/task-reference/create_media_buy"
+        schema_ref: 'media-buy/create-media-buy-request.json'
+        response_schema_ref: 'media-buy/create-media-buy-response.json'
+        doc_ref: '/media-buy/task-reference/create_media_buy'
         comply_scenario: deterministic_budget
         stateful: true
         context_outputs:
           - name: budget_media_buy_id
-            path: "media_buy_id"
+            path: 'media_buy_id'
         expected: |
           Create a media buy and return a media_buy_id.
 
         sample_request:
           account:
             brand:
-              domain: "acmeoutdoor.com"
-            operator: "pinnacle-agency.com"
+              domain: 'acmeoutdoor.com'
+            operator: 'pinnacle-agency.com'
           brand:
-            domain: "acmeoutdoor.com"
-          start_time: "2026-04-01T00:00:00Z"
-          end_time: "2026-06-30T23:59:59Z"
+            domain: 'acmeoutdoor.com'
+          start_time: '2026-04-01T00:00:00Z'
+          end_time: '2026-06-30T23:59:59Z'
           packages:
-            - product_id: "test-product"
+            - product_id: 'test-product'
               budget: 10000
-              pricing_option_id: "test-pricing"
+              pricing_option_id: 'test-pricing'
 
         validations:
           - check: response_schema
-            description: "Response matches create-media-buy-response.json schema"
+            description: 'Response matches create-media-buy-response.json schema'
 
       - id: simulate_budget_95
-        title: "Simulate 95% budget spend"
+        title: 'Simulate 95% budget spend'
         requires_tool: comply_test_controller
         narrative: |
           Simulate 95% of the media buy budget being spent. The agent should
@@ -887,20 +913,20 @@ phases:
           Return success with budget spend simulated to 95%.
 
         sample_request:
-          scenario: "simulate_budget_spend"
+          scenario: 'simulate_budget_spend'
           params:
-            media_buy_id: "{{budget_media_buy_id}}"
+            media_buy_id: '$context.budget_media_buy_id'
             spend_percentage: 95
 
         validations:
           - check: field_value
-            path: "success"
+            path: 'success'
             allowed_values:
               - true
-            description: "95% budget simulation succeeds"
+            description: '95% budget simulation succeeds'
 
       - id: simulate_budget_100
-        title: "Simulate 100% budget depletion"
+        title: 'Simulate 100% budget depletion'
         requires_tool: comply_test_controller
         narrative: |
           Simulate complete budget depletion. The agent should handle the fully
@@ -912,14 +938,14 @@ phases:
           Return success with budget fully depleted.
 
         sample_request:
-          scenario: "simulate_budget_spend"
+          scenario: 'simulate_budget_spend'
           params:
-            media_buy_id: "{{budget_media_buy_id}}"
+            media_buy_id: '$context.budget_media_buy_id'
             spend_percentage: 100
 
         validations:
           - check: field_value
-            path: "success"
+            path: 'success'
             allowed_values:
               - true
-            description: "100% budget depletion simulation succeeds"
+            description: '100% budget depletion simulation succeeds'