[EX-2672] Add campaign experiment lifecycle tools

TOOLS.md

@@ -1,4 +1,4 @@
-# Available Iterable MCP Tools (109 tools)
+# Available Iterable MCP Tools (118 tools)
 
 **Legend:**
 - 🔒 = Requires enabling user PII access
@@ -50,11 +50,20 @@
 - **track_bulk_events** 🔒✏️✉️: Track multiple events in a single request for better performance
 - **track_event** 🔒✏️✉️: Track a custom event for a user
 
-## Experiments (4 tools)
+## Experiments (13 tools)
+- **cancel_experiment** ✏️: Cancel a running or winner_found campaign experiment without declaring a winner.
+- **copy_experiment_variant** ✏️: Copy a template into a new variant on a draft or ready campaign experiment.
+- **create_experiment** ✏️: Create a draft campaign experiment. The campaign template becomes the control.
+- **declare_experiment_winner** ✏️✉️: Declare a winning variant and end a running or winner_found campaign experiment.
+- **delete_experiment** ✏️: Delete a draft, ready, or errored campaign experiment.
 - **get_experiment**: Get detailed information about a specific experiment by ID, including variants summary and constraints
 - **get_experiment_metrics**: Get experiment metrics for A/B testing analysis (currently supports email experiments only)
+- **get_experiment_totals**: Get lifetime send and conversion totals for a campaign experiment. Holdout is omitted, missing counts are 0, and lift and confidence are not included.
+- **get_experiment_trends**: Get the performance time series for a campaign experiment. Provide both startDateTime and endDateTime, or omit both to use the run window. The span is at most 31 days; a longer default window is clipped to the last 31 days.
 - **get_experiment_variants**: Get variant content for an experiment, including subject lines, preheaders, HTML source, and plain text
 - **list_experiments**: List experiments with optional filtering by campaign, status, and date range. Supports pagination.
+- **start_experiment** ✏️✉️: Start a draft or ready campaign experiment.
+- **update_experiment_settings** ✏️: Update settings on a draft or ready campaign experiment. Omit a field to leave it unchanged.
 
 ## Journeys (2 tools)
 - **get_journeys**: Get journeys (workflows) with optional pagination and state filtering

package.json

@@ -76,7 +76,7 @@
   },
   "dependencies": {
     "@alcyone-labs/zod-to-json-schema": "4.0.10",
-    "@iterable/api": "0.12.0",
+    "@iterable/api": "0.13.0",
     "@modelcontextprotocol/sdk": "1.18.1",
     "@primno/dpapi": "2.0.1",
     "@types/json-schema": "7.0.15",

src/tool-filter.ts

@@ -13,13 +13,18 @@ export const NON_PII_TOOLS: Set<string> = new Set([
   "archive_campaigns",
   "bulk_delete_catalog_items",
   "cancel_campaign",
+  "cancel_experiment",
+  "copy_experiment_variant",
   "create_blast_campaign",
+  "create_experiment",
   "create_triggered_campaign",
   "create_catalog",
   "create_list",
   "create_snippet",
   "deactivate_triggered_campaign",
+  "declare_experiment_winner",
   "delete_catalog",
+  "delete_experiment",
   "delete_catalog_item",
   "delete_list",
   "delete_snippet",
@@ -36,6 +41,8 @@ export const NON_PII_TOOLS: Set<string> = new Set([
   "get_email_template",
   "get_experiment",
   "get_experiment_metrics",
+  "get_experiment_totals",
+  "get_experiment_trends",
   "get_experiment_variants",
   "get_inapp_template",
   "get_journeys",
@@ -57,11 +64,13 @@ export const NON_PII_TOOLS: Set<string> = new Set([
   "replace_catalog_item",
   "schedule_campaign",
   "send_campaign",
+  "start_experiment",
   "trigger_campaign",
   "update_catalog_field_mappings",
   "partial_update_catalog_items",
   "replace_catalog_items",
   "update_email_template",
+  "update_experiment_settings",
   "update_inapp_template",
   "update_push_template",
   "update_sms_template",
@@ -87,6 +96,8 @@ export const READ_ONLY_TOOLS: Set<string> = new Set([
   "get_embedded_messages",
   "get_experiment",
   "get_experiment_metrics",
+  "get_experiment_totals",
+  "get_experiment_trends",
   "get_experiment_variants",
   "get_export_files",
   "get_export_jobs",
@@ -125,6 +136,8 @@ export const SEND_TOOLS: Set<string> = new Set([
   "send_campaign",
   "trigger_campaign",
   "schedule_campaign",
+  "start_experiment",
+  "declare_experiment_winner",
   // Triggered campaigns can cause sends upon activation; block unless explicitly allowed
   "activate_triggered_campaign",
   // Journey triggers enqueue users which may send

src/tools/experiments.ts

@@ -4,10 +4,17 @@
 
 import type { IterableClient } from "@iterable/api";
 import {
+  CopyExperimentVariantParamsSchema,
+  CreateExperimentParamsSchema,
+  DeclareExperimentWinnerParamsSchema,
+  ExperimentIdParamsSchema,
   GetExperimentMetricsParamsSchema,
   GetExperimentParamsSchema,
+  GetExperimentTotalsParamsSchema,
+  GetExperimentTrendsParamsSchema,
   GetExperimentVariantsParamsSchema,
   ListExperimentsParamsSchema,
+  UpdateExperimentSettingsParamsSchema,
 } from "@iterable/api";
 import type { Tool } from "@modelcontextprotocol/sdk/types.js";
 
@@ -43,5 +50,66 @@ export function createExperimentTools(client: IterableClient): Tool[] {
       schema: GetExperimentMetricsParamsSchema,
       execute: (params) => client.getExperimentMetrics(params),
     }),
+    createTool({
+      name: "get_experiment_totals",
+      description:
+        "Get lifetime send and conversion totals for a campaign experiment. Holdout is omitted, missing counts are 0, and lift and confidence are not included.",
+      schema: GetExperimentTotalsParamsSchema,
+      execute: (params) => client.getExperimentTotals(params),
+    }),
+    createTool({
+      name: "get_experiment_trends",
+      description:
+        "Get the performance time series for a campaign experiment. Provide both startDateTime and endDateTime, or omit both to use the run window. The span is at most 31 days; a longer default window is clipped to the last 31 days.",
+      schema: GetExperimentTrendsParamsSchema,
+      execute: (params) => client.getExperimentTrends(params),
+    }),
+    createTool({
+      name: "create_experiment",
+      description:
+        "Create a draft campaign experiment. The campaign template becomes the control.",
+      schema: CreateExperimentParamsSchema,
+      execute: (params) => client.createExperiment(params),
+    }),
+    createTool({
+      name: "copy_experiment_variant",
+      description:
+        "Copy a template into a new variant on a draft or ready campaign experiment.",
+      schema: CopyExperimentVariantParamsSchema,
+      execute: (params) => client.copyExperimentVariant(params),
+    }),
+    createTool({
+      name: "update_experiment_settings",
+      description:
+        "Update settings on a draft or ready campaign experiment. Omit a field to leave it unchanged.",
+      schema: UpdateExperimentSettingsParamsSchema,
+      execute: (params) => client.updateExperimentSettings(params),
+    }),
+    createTool({
+      name: "start_experiment",
+      description: "Start a draft or ready campaign experiment.",
+      schema: ExperimentIdParamsSchema,
+      execute: (params) => client.startExperiment(params),
+    }),
+    createTool({
+      name: "cancel_experiment",
+      description:
+        "Cancel a running or winner_found campaign experiment without declaring a winner.",
+      schema: ExperimentIdParamsSchema,
+      execute: (params) => client.cancelExperiment(params),
+    }),
+    createTool({
+      name: "declare_experiment_winner",
+      description:
+        "Declare a winning variant and end a running or winner_found campaign experiment.",
+      schema: DeclareExperimentWinnerParamsSchema,
+      execute: (params) => client.declareExperimentWinner(params),
+    }),
+    createTool({
+      name: "delete_experiment",
+      description: "Delete a draft, ready, or errored campaign experiment.",
+      schema: ExperimentIdParamsSchema,
+      execute: (params) => client.deleteExperiment(params),
+    }),
   ];
 }