--- title: System tools for Optimizely Data Platform (ODP) slug: agent-platform/docs/system-tools-for-optimizely-data-platform-odp docTags: createdAt: 2026-08-12T15:37:08.000Z --- System tools are built-in features that help Mark take action. Each tool performs a specific task, such as creating a campaign, uploading files, or generating images. Think of tools like attachments on a Swiss Army knife. Each one has a distinct purpose that helps you get work done. In addition to the [system tools available in Agent Platform](docId\:ZExsvHuX-MENKqR8_eAvw), Optimizely Data Platform (ODP) includes a set of system tools designed to support getting campaign information. :::hint{type="info"} **Note** Ask Mark what tools it has at any time. For example, enter "*Please list the tools you have with a brief description of what they do and the parameters*" into [Mark chat](../../use-opal/opal-chat/access-and-explore-opal-chat.mdx). ::: Click a tool's name to expand it and learn when to use it, its required and optional parameters, and example prompts for calling the tool. If you do not provide a required parameter, Mark prompts you for it. :::ExpandableHeading ### `create_odp_audience` – Creates a new audience (segment) in ODP, supporting both standard and realtime types. **When to use** - Create a new realtime audience in ODP with a custom definition and ID. - Create a new standard audience segment in ODP with a display name and definition. - Build a new audience segment from a JSON definition to target specific customer behaviors. - Configure a new ODP segment as part of a personalization or targeting strategy. **Parameters** - `type` – Type of audience: `standard` or `realtime`. - `definition` – Definition of the audience in JSON format. - (Optional) `id` – ID of the realtime audience to be created (required for `realtime`). - (Optional) `display_name` – Display name of the audience (required for `standard`). **Example prompts** - *Create a realtime ODP audience with id "high\_value\_customers" and a definition that targets customers with lifetime value over 500.* - *I need a new standard audience in ODP called "Newsletter Subscribers" with a definition filtering by email opt-in status.* - *Can you create a realtime ODP segment with id "recent\_purchasers" and display name "Recent Purchasers" that includes customers who purchased in the last 30 days?* ::: :::ExpandableHeading ### `create_odp_audiences` – Creates a new real-time segment in ODP based on a natural language description. **When to use** - Quickly create a real-time ODP audience from a plain-language description without writing JSON. - Generate one real-time segment at a time using a conversational prompt. - Build a targeted audience segment when you know the criteria but do not want to manually construct the definition. **Parameters** - `user_query` – A natural language query of a segment or audience to create in ODP. Use the exact query that the user wrote without rewriting it. The query will be normalized (lowercased, punctuation removed, and extra spaces cleaned). **Example prompts** - *Create an ODP segment for customers who abandoned their cart in the last 7 days.* - *Build me an ODP audience of users who have made more than 3 purchases in the past month.* - *I need a real-time ODP segment for customers located in the United States who opened an email in the last 14 days.* ::: :::ExpandableHeading ### `fetch_schema` – Retrieves the complete ODP data model schema for the connected account. **When to use** - Understand what customer data is available in ODP, such as email, loyalty tier, or purchase history. - Explore which event types are being tracked or inspect the event schema structure. - Identify available fields before building audience segments or queries. - Understand data relationships, for example how customers relate to orders or events. - Answer questions like "what data do we have," "what fields are available," "show me the schema," or "what customer attributes exist." **Parameters** - (Optional) `dimensions` – Optional comma-separated list of specific dimension or object names to retrieve (for example, `customers,events,products`). Dimension names must be in lowercase, snake\_case, and plural form (for example, `customers` not `Customer`). When provided, Mark only returns schema information for the specified dimensions. When omitted or empty, Mark returns the complete schema for all dimensions in the account. **Example prompts** - *Show me the full ODP data model schema.* - *What customer attributes are available in ODP?* - *Fetch the ODP schema for customers and events only.* - *What fields are available in the orders dimension in ODP?* ::: :::ExpandableHeading ### `get_odp_audience_by_id` – Retrieves details for a specific ODP segment (audience) by its ID. **When to use** - Look up a specific ODP audience or segment by its unique identifier. - Inspect the definition, configuration, or metadata of an existing segment. - Verify the details of a realtime or standard audience before updating or referencing it. **Parameters** - `id` – ID of segment to be returned. - `type` – Type of segment: `standard` or `realtime`. **Example prompts** - *Show me the details of the ODP realtime segment with id "high\_value\_customers".* - *Get the definition of standard ODP audience "12345".* - *Look up the ODP segment "recent\_purchasers" — it's a realtime audience.* ::: :::ExpandableHeading ### `get_odp_audience_list` – Retrieves a list of all segments (audiences) in ODP by type. **When to use** - List all available realtime or standard segments in ODP. - Get an overview of existing ODP audiences and their descriptions. - Discover which segments are configured before creating or updating one. **Parameters** - `type` – Type of segment: `standard` or `realtime`. **Example prompts** - *Show me all realtime ODP segments.* - *List all standard audiences in ODP.* - *What ODP realtime audiences do we have configured?* ::: :::ExpandableHeading ### `odp_create_custom_fields` – Adds one or more custom fields to an existing ODP dimension. **When to use** - Add fields to an existing dimension, such as `customers`, `events`, or `orders`. - Create several fields on the same dimension in a single request. - Add a field that contains sensitive data and mark it as personally identifiable information (PII) so ODP applies the correct access controls. - Expose a field through the public API key by enabling public read access. **Parameters** - `object` – The dimension or object name to add fields to (for example, `customers`, `events`). Must be lowercase, snake\_case, plural form. - `fields` – JSON array of field definitions. - Each field object accepts the following: - `name` – Field name in snake\_case. - `type` – One of `string`, `number`, `boolean`, `timestamp` - Each field also accepts the following optional properties: - `display_name` – Human-readable label, auto-generated from `name` if omitted. - `description` – Field description string. - `public_read` – Enable access through the public API key, defaults to `false`. - `access_policies` – PII access control, an object with a PII property containing `enable` and `suggestible`). **Example prompts** - *Add a* `preferred_language` *string field to the customers dimension in ODP.* - *Create fields* `loyalty_tier` *(string) and* `loyalty_points` *(number) on the customers object.* - *Add a field called* `ssn` *of type string to customers and mark it as PII.* - *Add a boolean field* `email_opt_in` *to customers and make it readable through the public API key.* ::: :::ExpandableHeading ### `odp_create_dimension` – Creates a custom dimension (object) in ODP. **When to use** - Add a custom object to your ODP data model to represent a business entity that does not fit an existing dimension, such as Household or Loyalty Program. - Define a dimension and its primary key before you add fields or relationships to it. - Extend your ODP schema to support a data source or use case that existing dimensions do not cover. - Model a physical or organizational entity, such as stores or franchises, that you plan to join to customers or events. **Parameters** - `display_name` – Human-readable label for the dimension (for example, *Household*). - `name` – API identifier for the dimension. Lowercase letters, numbers, and underscores only. Must start with a letter. Maximum 255 characters (for example, `household`). - `primary_field_display_name` – Human-readable label for the dimension's primary key (for example, *Household ID*). - `primary_field_name` – API identifier for the primary key. Follows the same format rules as `name` (for example, `household_id`). - `primary_field_type` – Data type of the primary key: `string` or `number`. - (Optional) `description` – Free-text description of the dimension. **Example prompts** - *Create an ODP dimension called "Household" with a string primary key "Household ID."* - *Add a "Loyalty Program" object to ODP with a numeric primary key called "Program ID."* - *Configure a custom dimension "Stores" in ODP with a string primary field* `store_id`*.* - *Create a "Subscriptions" dimension in ODP with a string primary key* `subscription_id` *and a description explaining that it tracks recurring plans.* ::: ::::ExpandableHeading ### `odp_create_field_link` – Creates a field link so an events field value copies continuously to a customer profile field. **When to use** - Sync a value from an events field to the matching customer profile field automatically. - Persist an event value for longer than RTS (Real-time audiences)'s max 28 days window. - Configure ongoing data flow from events into customer attributes without manual updates. - Surface a behavioral signal on the customer profile so segments and campaigns target it. **Parameters** - `source_field` – The field name on the events dimension to copy from (for example, `total_revenue`). - `destination_field` – The field name on the customers dimension to copy to (for example, `lifetime_revenue`). **Example prompts** - *Link the* `total_revenue` *event field to the* `lifetime_revenue` *customer field in ODP.* - *Create a field link so* `last_login_device` *from events updates the customers* `last_device` *field.* - *Configure a continuous copy from* `events.email_opt_in` *to* `customers.email_opt_in`*.* - *Copy the* `last_order_date` *event field onto the customer profile in ODP.* :::hint{type="info"} **Note** - Both fields must exist in the schema and have compatible data types before you create a link. - Only events received after you create the link sync to the profile. If a link request fails, review the tool's response for details. ::: :::: :::ExpandableHeading ### `odp_create_identifier` – Creates a custom identifier in ODP for customer identification and merging. **When to use** - Register a type of customer identifier, such as a third-party account ID, so ODP matches and merges profiles on it. - Add an identifier that campaigns use to message customers, such as a push token. - Configure an identifier whose values do not map to a single person, such as a shared device token. - Support a new integration that sends its own identifier values into ODP. **Parameters** - `label` – Human-readable identifier label, without the suffix (for example, *Platform*, not *Platform ID*). Used to derive both the display name and the API name. - `suffix` – Identifier type: `id`, `hash`, `number`, `token`, `alias`, `address`, `key`, or `push_token`. - `merge_confidence` – `high` (unique per person) or `low` (shared across people). - (Optional) `description` – Short description of the identifier. - (Optional) `messaging` – Whether campaigns use the identifier to message customers. Defaults to true when `suffix` is `push_token`. - (Optional) `public_read` – Whether the public API key reads the identifier. Never enable this for PII. Defaults to `false`. **Example prompts** - *Create a high-confidence identifier in ODP for "Shopify" with suffix* `id`*.* - *Add a low-confidence "Device" identifier of type token in ODP.* - *Configure a push token identifier called "Mobile App" that campaigns use for messaging.* - *Create a "Loyalty" identifier in ODP with suffix* `number` *and high merge confidence.* ::: :::ExpandableHeading ### `odp_create_relationship` – Creates a parent-child relationship between two dimensions in ODP. **When to use** - Link two existing dimensions, such as customers and products, through a parent-child relationship. - Define how a field on one dimension references records in another dimension. - Configure the join that queries and reports need to span related objects. - Connect a custom dimension you created to a standard dimension, such as customers. **Parameters** - `display_name` – Human-readable name of the relationship, typically the parent field name (for example, *First Product*). - `name` – API name. Lowercase letters, numbers, and underscores only. Must start with a letter (for example, `first_product`). - `parent_object_name` – The dimension that contains the field pointing to another dimension (for example, `customers`). - `child_object` – The dimension referenced by the parent's join field (for example, `products`). - `join_fields` – JSON array of `{parent, child}` field-name pairs that describe how the two dimensions join. **Example prompts** - *Create a relationship in ODP called "First Product" that joins* `customers.first_product_id` *to* `products.product_id`*.* - *Link the orders and customers dimensions using* `order.customer_id` *and* `customers.customer_id`*.* - *Configure a parent-child relationship named* `primary_store` *between customers and stores.* - *Join my custom households dimension to customers on* `household_id`*.* ::: :::ExpandableHeading ### `odp_query_event_logs` – Queries event log data from ODP and returns matching records. **When to use** - Retrieve recent event activity for a specific customer or set of customers in ODP. - Investigate what events were tracked for a particular user, such as purchases, page views, or email interactions. - Query ODP event logs to troubleshoot or validate that events are being recorded correctly. - Analyze customer behavior patterns by querying event data with filters. **Parameters** - `queries` – The natural language query describing what event log data to retrieve from ODP. **Example prompts** - *Show me the last 10 events for the customer with email "jane\@example.com" in ODP.* - *Query ODP event logs for all purchase events in the last 24 hours.* - *What events have been tracked for customer ID "abc123" in ODP?* ::: :::ExpandableHeading ### `suggest_odp_audiences` – Suggests three real-time ODP audience segments based on a natural language description. **When to use** - Get audience segment suggestions from ODP before committing to a specific definition. - Explore possible audience configurations when you have a general targeting idea but want recommendations. - Generate multiple audience segment options to compare before creating one. **Parameters** - None. **Example prompts** - *Suggest ODP audiences for targeting high-value customers who have purchased more than three times.* - *What ODP audience segments would you recommend for re-engaging users who have not visited in 30 days?* - *Suggest some ODP segments for customers interested in electronics based on their browsing behavior.* ::: :::ExpandableHeading ### `update_odp_audience` – Updates an existing audience (segment) in ODP. **When to use** - Modify the definition of an existing ODP audience segment. - Change the display name of a standard audience in ODP. - Update the targeting criteria for a realtime or standard segment that already exists. **Parameters** - `id` – Id of the audience to be updated. - `type` – Type of audience: `standard` or `realtime`. - `definition` – The segment definition in JSON format. - (Optional) `description` – Description for the segment (required for realtime). - (Optional) `display_name` – Display name of the audience (required for standard). **Example prompts** - *Update the ODP realtime segment "high\_value\_customers" to include customers with lifetime value over 1000 instead of 500.* - *Change the display name of standard ODP audience "12345" to "VIP Customers".* - *Update the definition of the ODP realtime audience "recent\_purchasers" to target purchases in the last 14 days instead of 30.* ::: :::ExpandableHeading ### `validate_odp_audience` – Validates an ODP audience definition without creating or updating it. **When to use** - Check whether an audience definition is valid before creating or updating a segment in ODP. - Troubleshoot errors in an ODP audience definition by validating its structure. - Confirm that a JSON audience definition conforms to ODP requirements before applying it. **Parameters** - `definition` – Definition of the audience in JSON format to validate. - `type` – Type of audience: `standard` or `realtime`. **Example prompts** - *Validate this ODP audience definition before I create the segment: \{"filter": \[\{"field": "lifetime\_value", "operator": "gt", "value": 500}]}.* - *Check if this realtime ODP audience definition is valid: \{"event": "purchase", "days": 30}.* - *Can you validate my standard ODP audience definition to make sure it has no errors?* ::: :::hint{type="info"} **Note** If you use Opti ID, administrators can turn off generative AI in the Opti ID Admin Center. See [Turn generative AI off across Optimizely applications](docId\:X5U0kKD2azV5cIdNqwGNj). :::