@salesforce/afv-skills 1.65.0 → 1.66.0

package/skills/health-cloud-care-provider-search-config-configure/SKILL.md

@@ -1,134 +0,0 @@
----
-name: health-cloud-care-provider-search-config-configure
-description: "Use this skill when users want to manage Care Provider Search Config records (CareProviderSearchConfig) with standard sObject REST (/services/data/v67.0/sobjects/CareProviderSearchConfig). Trigger when users mention care, provider, search, config, care provider search config."
-metadata:
-  version: "1.0"
----
-
-# Managing Care Provider Search Config
-
-## When to Use This Skill
-
-Manage Care Provider Search Config records (CareProviderSearchConfig) with standard sObject REST (/services/data/v67.0/sobjects/CareProviderSearchConfig). Each record maps one source field on Healthcare Provider or Healthcare Practitioner Facility onto a custom searchable field on Care Provider Searchable Field, so Provider Search can filter on it and show it on result cards. It is optional and additive: with no records, Provider Search still indexes the standard provider fields. Required to create: MasterLabel, DeveloperName (unique), SourceField and TargetField (API names). MappedObject defaults to HealthcareProvider; set it to HealthcarePractitionerFacility for a facility field, and the source field must belong to the mapped object. IsActive defaults to false, so send IsActive=true to use the mapping. The mapping rules are checked on every create and update, even when the record is inactive: the target must be a custom Text or Text Area field at least as long as the source and cannot be a standard field or another type, a Text Area target is allowed only for a multi-select picklist source, and a target can have only one active mapping (deactivate the old record to free it). A new mapping takes effect on a later Provider Search sync, and the field must also be added to the Provider Search page in the Lightning page builder.
-
-## Workflow
-
-Create the target custom field on Care Provider Searchable Field first (see the Provider Search Settings skill, step ensure-searchable-field-custom-field), then create the config record. Common errors and their causes: "This field isn't valid for this context" means the source or target field does not exist on the mapped object, or SourceField/TargetField was left out; "The target field's length must be at least equal to that of the source field"; "The target field cannot be a standard field"; "The provider search cannot be configured. The target field type needs to be text or text area"; "The source field type must be picklist (multi-select) if the target field is type text area"; "This target field already has a source field" means another active record uses that target. To move a target to a different source, set the old record IsActive to false (or delete it) first. This skill is optional for Provider Search; do not report Provider Search as broken just because no config record exists.
-
-## Critical Constraints
-
-**Preconditions:**
-
-- Provider Search must be enabled in the org and the user must have Provider Search access. (check: `Provider Search enabled`)
-- The user needs the Customize Application permission to create, edit or delete a config record.
-- The target custom field (Text or Text Area) must already exist on Care Provider Searchable Field, and the source field must exist on the mapped object (Healthcare Provider or Healthcare Practitioner Facility).
-
-## Operations Reference
-
-Operations grouped by purpose. Use these as the building blocks for the workflows above.
-
-### Summary
-
-| Operation | Purpose | Status | Call | Depends on |
-|-----------|---------|--------|------|------------|
-| `describe` | read | — | `GET /services/data/v67.0/sobjects/CareProviderSearchConfig/describe` | — |
-| `list` | read | — | `GET /services/data/v67.0/query` | — |
-| `get-by-id` | read | — | `GET /services/data/v67.0/sobjects/CareProviderSearchConfig/{Id}` | `list` |
-| `create` | write | — | `POST /services/data/v67.0/sobjects/CareProviderSearchConfig` | `describe` |
-| `update` | write | — | `PATCH /services/data/v67.0/sobjects/CareProviderSearchConfig/{Id}` | `get-by-id` |
-| `delete` | write | — | `DELETE /services/data/v67.0/sobjects/CareProviderSearchConfig/{Id}` | `get-by-id` |
-
-### Dependency graph
-
-```mermaid
-graph TD
-  describe["describe (read)"]
-  list["list (read)"]
-  get_by_id["get-by-id (read)"]
-  create["create (write)"]
-  update["update (write)"]
-  delete["delete (write)"]
-  list --> get_by_id
-  describe --> create
-  get_by_id --> update
-  get_by_id --> delete
-```
-
-### Read operations
-
-#### `describe`
-
-Retrieve the CareProviderSearchConfig sObject describe — fields, picklist values, FLS, createable/updateable flags.
-
-**Call:** `GET /services/data/v67.0/sobjects/CareProviderSearchConfig/describe`
-
-**Output:** DescribeSObjectResult for CareProviderSearchConfig — use this to build the field list for list / create / update payloads.
-
-#### `list`
-
-List CareProviderSearchConfig records. Pair this with /describe first to pick field names.
-
-**Call:** `GET /services/data/v67.0/query`
-
-**Inputs:**
-
-- `where` *(`String`)* — Optional SOQL WHERE clause to filter Care Provider Search Config records.
-- `limit` *(`int`)* — Optional SOQL LIMIT value.
-- `q` *(`string`)*
-
-**Output:** Query response with records[] of CareProviderSearchConfig — Id plus any SELECTed fields.
-
-#### `get-by-id`
-
-Retrieve a specific CareProviderSearchConfig record by Id.
-
-**Call:** `GET /services/data/v67.0/sobjects/CareProviderSearchConfig/{Id}`
-
-**Inputs:**
-
-- `Id` *(`ID`)* — Id of the CareProviderSearchConfig record to fetch.
-
-**Depends on:** `list`
-
-**Output:** Single CareProviderSearchConfig record with all readable fields per user FLS.
-
-### Write operations
-
-#### `create`
-
-Create a Care Provider Search Config record mapping one source field to one custom searchable field. Required: MasterLabel, DeveloperName, SourceField, TargetField. MappedObject defaults to HealthcareProvider (or set HealthcarePractitionerFacility); IsActive defaults to false, so send IsActive=true to use it. Fails if either field does not exist on the mapped object, the target is a standard field or not a Text/Text Area field at least as long as the source, a Text Area target is used with a source that is not a multi-select picklist, or the target already belongs to another active record.
-
-**Call:** `POST /services/data/v67.0/sobjects/CareProviderSearchConfig`
-
-**Inputs:**
-
-- `body` *(`object`)* — Field/value map for the new CareProviderSearchConfig. Populate from /describe's createableFields list.
-
-**Depends on:** `describe`
-
-**Rollback:** Call delete with the returned Id to undo the create.
-
-#### `update`
-
-Update an existing CareProviderSearchConfig record. Merge-patch semantics.
-
-**Call:** `PATCH /services/data/v67.0/sobjects/CareProviderSearchConfig/{Id}`
-
-**Inputs:**
-
-- `Id` *(`ID`)* — Id of the CareProviderSearchConfig to update.
-- `body` *(`object`)* — Partial field/value map — only include fields you want to change.
-
-**Depends on:** `get-by-id`
-
-#### `delete`
-
-Delete a CareProviderSearchConfig record.
-
-**Call:** `DELETE /services/data/v67.0/sobjects/CareProviderSearchConfig/{Id}`
-
-**Inputs:**
-
-- `Id` *(`ID`)* — Id of the CareProviderSearchConfig to delete.
-
-**Depends on:** `get-by-id`

package/skills/health-cloud-care-requests-create/SKILL.md

@@ -1,178 +0,0 @@
----
-name: health-cloud-care-requests-create
-description: "Use this skill when users want to create, copy, update, and read Health Cloud care requests via the Care Services Connect API instead of the UI. Also use when a Health Cloud user wants to create/duplicate/update a utilization-management care request — for example an authorization request, an appeal, or a care request of type Referral. Trigger when users mention health, cloud, services, provider, search, settings, provider search settings."
-metadata:
-  version: "1.0"
-  accessCheck: "UserPermissions.HealthCloudUtilizationMgmtUser"
----
-
-# Managing Health Cloud Services
-
-## When to Use This Skill
-
-Create, copy, update, and read Health Cloud care requests via the Care Services Connect API instead of the UI. Care requests are utilization-management records — authorizations, appeals, admissions, and a Referral request type among them: "create" posts a new CareRequest through POST /connect/health/care-services; "copy" reads an existing record with GET /connect/health/care-services/{ids} and re-creates it as a new CareRequest (there is no server-side copy endpoint — copy is a client-side read-then-create); "update" PATCHes the collection path. The request body is a top-level careRequestsList array; each item carries a careRequest sub-object (the record's own fields) plus optional related sections. A create requires name plus a request/record type on careRequest, and the request type must map to a CareRequest RecordType configured on the org (the requestType picklist is Admission, Appeal, Complaint, Concurrent Review, Drug Request, Grievance, Referral, Service Request). Use when a Health Cloud user wants to create/duplicate/update a utilization-management care request — for example an authorization request, an appeal, or a care request of type Referral. This is the utilization-management CareRequest object; it is NOT the Provider Search "refer to a provider" clinical-referral flow, which creates a ClinicalServiceRequest via POST /connect/health/referral-management/referrals — use that flow (not this skill) to refer a member to a provider found in Provider Search. Requires the HealthCloudUtilizationMgmtUser permission; care-services is available from API v46.0.
-
-## Workflow
-
-Create a care request / referral: create-care-request (POST /connect/health/care-services with a top-level careRequestsList array) → verify-create-care-request (GET the new id back). Copy a record: read-care-request (GET the source with a broad sectionNames list) → copy-care-request (POST the mapped sections as a new CareRequest, dropping the source id) → verify-copy-care-request (confirm the copied id differs from the source). Update: update-care-request (PATCH the collection path with the same careRequestsList array, id on the careRequest sub-object) → verify-update-care-request. Read only: read-care-request.
-
-Contract truths:
-- The backing object is the custom object CareRequest, NOT a standard Referral sObject. Create/copy/update/read all go through the Care Services Connect API, not sObject CRUD on a "Referral". A care request of type Referral is just a care-request operation with careRequest.requestType (or a Referral RecordType) set. This is NOT the Provider Search "refer to a provider" clinical-referral flow — that flow creates a ClinicalServiceRequest via POST /connect/health/referral-management/referrals, a different object and endpoint; use it (not this skill) to refer a member to a provider found in Provider Search.
-- The requestType (Care Request Type) picklist values are Admission, Appeal, Complaint, Concurrent Review, Drug Request, Grievance, Referral, and Service Request (defined on CareRequestConfiguration.CareRequestType). Supply one of these as careRequest.requestType, OR supply careRequest.recordType directly as a matching Case+CareRequest RecordType developer name.
-- The request body is a top-level careRequestsList array — there is NO createCareAuthData/updateCareAuthData wrapper. Both POST (create) and PATCH (update) use the identical top-level shape { "careRequestsList": [ ... ] } and the same collection path /connect/health/care-services (no id in the URL). For an update, the target record id lives on the careRequest sub-object of the list item, not in the URL.
-- Each list item wraps the record's own fields in a careRequest sub-object and carries optional sibling sections (subscriber, patient, careDiagnoses, careRequestItems, careRequestDrugs, transportLocations, careRequestCase). A create requires careRequest.name and a request/record type (careRequest.requestType or careRequest.recordType); an empty item is rejected ("Specify a valid value for either request type or record type").
-- The request type must map to a CareRequest RecordType configured on the org, or the create is rejected ("Specify a valid record type for the request type"). The resolved record-type developer name must exist as a RecordType on BOTH the Case and the CareRequest objects with the identical developer name (a care request creates a backing Case), and must be visible to the running user's profile — otherwise the create fails "Invalid record type in this care request." or "this ID value isn't valid for the user." Ensure the matching Case+CareRequest record-type pair exists and is assigned before creating (see preconditions).
-- Record type must be supplied on UPDATE as well as create. A PATCH item carrying only an id and changed fields is rejected ("Invalid record type in this care request.") because the update re-resolves the record type. The published input schema marks recordType optional, but the service re-resolves it at runtime on every call, so in practice always include careRequest.recordType (or requestType) on the update item alongside the id.
-- "Copy" has no server endpoint: it is a client-side GET-source → POST-new. Drop server-owned id fields from the mapped body so a new CareRequest is created instead of updating the source; keep the request/record type so the copy still resolves to a valid RecordType.
-- The response is CareAuthorizationRequestResultRepresentation {code, message, isSuccess, result[]}; each result carries id, created, status, objectType, relatedRecords. Read new ids from result[].id and check created to tell create (true) from update (false).
-- Requires the HealthCloudUtilizationMgmtUser permission; the care-services resource is available from v46.0.
-
-## Critical Constraints
-
-**Preconditions:**
-
-- User must hold the Health Cloud Utilization Management user permission to call the Care Services API. (check: `UserPermissions.HealthCloudUtilizationMgmtUser`)
-- The Care Services Connect API is available from API version 46.0 (v46.0+). Call it at v46.0 or later; older versions do not expose the care-services resource. (check: `API version >= v46.0`)
-- At least one active CareRequest RecordType must be configured on the org before a create can succeed. A create supplies a request type (or record type) on the careRequest sub-object, and the API validates that the request type maps to an existing CareRequest RecordType — a create against an org with no matching RecordType is rejected ("Specify a valid record type for the request type"). Health Cloud utilization-management record types come as matching pairs — the resolved record-type developer name must exist as a RecordType on BOTH the Case and the CareRequest objects with the identical developer name, because a care request creates a backing Case. If only the CareRequest-side record type exists, the create still fails ("Invalid record type in this care request."), and the running user must have that record type visible on their profile (an unassigned record type is rejected with "this ID value isn't valid for the user"). This is org data configuration, not an API-contract input. (check: `an active RecordType with the same developer name exists on BOTH the CareRequest and the Case objects, visible to the running user, matching the request type used`)
-- A create that identifies the member (the patient section, via patient.memberId or patient.firstName + patient.lastName) resolves that member by querying MemberPlan on the member's name — and because Member is an Account reference, the lookup reads Member.FirstName / Member.LastName. Those fields exist only when Person Accounts is enabled on the org; a member represented as a business Account has no first/last name and cannot be resolved, so the create is rejected ("Enter the member ID, or the member's first and last names"). Identifying a member BY NAME therefore requires Person Accounts enabled on the org and, for that name-based lookup, an enrolled member the name resolves to (a Person Account with a MemberPlan); identifying a member by patient.memberId instead passes the member's Account Id directly and does not depend on a name lookup. The patient section itself is optional in the published input schema, but once a create supplies member identity it must resolve, and on a non–Person-Account org a name-identified member cannot resolve. This is org data configuration, not an API-contract input. (check: `Person Accounts is enabled on the org; a member identified by name resolves via a MemberPlan-enrolled Person Account, while a member identified by patient.memberId is passed as the member's Account Id directly`)
-
-## Verification Checklist
-
-After every write operation, confirm the effect by re-reading state:
-
-- After `create-care-request`: call `verify-create-care-request`
-- After `copy-care-request`: call `verify-copy-care-request`
-- After `update-care-request`: call `verify-update-care-request`
-
-## Operations Reference
-
-Operations grouped by purpose. Use these as the building blocks for the workflows above.
-
-### Summary
-
-| Operation | Purpose | Status | Call | Depends on |
-|-----------|---------|--------|------|------------|
-| `create-care-request` | write | implemented | `POST /connect/health/care-services` | — |
-| `verify-create-care-request` | verify | — | `getCareServices` | `create-care-request` |
-| `read-care-request` | read | — | `getCareServices` | — |
-| `copy-care-request` | write | implemented | `postCareServices` | `read-care-request` |
-| `verify-copy-care-request` | verify | — | `getCareServices` | `copy-care-request` |
-| `update-care-request` | write | implemented | `patchCareServices` | — |
-| `verify-update-care-request` | verify | — | `getCareServices` | `update-care-request` |
-
-### Dependency graph
-
-```mermaid
-graph TD
-  create_care_request["create-care-request (write)"]
-  verify_create_care_request["verify-create-care-request (verify)"]
-  read_care_request["read-care-request (read)"]
-  copy_care_request["copy-care-request (write)"]
-  verify_copy_care_request["verify-copy-care-request (verify)"]
-  update_care_request["update-care-request (write)"]
-  verify_update_care_request["verify-update-care-request (verify)"]
-  create_care_request --> verify_create_care_request
-  read_care_request --> copy_care_request
-  copy_care_request --> verify_copy_care_request
-  update_care_request --> verify_update_care_request
-```
-
-### Read operations
-
-#### `read-care-request`
-
-Read a care request in full — for direct inspection or as the source of a copy. Care Services exposes no server-side copy/clone operation, so a copy is a client-side read-then-create and this step is its read half: GET the CareRequest with idName=Id and a broad sectionNames list (Subscriber;Patient;CareDiagnoses;CareRequestItems;CareRequestDrugs;TransportLocations;CareRequest) so every section that should carry over is returned. The copy step maps this payload into a fresh create body.
-
-**Call:** `getCareServices`
-
-**Inputs:**
-
-- `ids` *(`String`)* — The CareRequest id to read (the source id when copying).
-- `idName` *(`string`)*
-- `sectionNames` *(`string`)*
-
-**Output:** Full CareRequest with the requested related sections; when copying, the fields to carry into the new record.
-
-### Write operations
-
-#### `create-care-request`
-
-Create a care request (for a referral, supply the referral's fields). The body is a top-level careRequestsList array; each list item is a CareAuthorizationRequestInputRepresentation holding a careRequest sub-object (the record's own fields) plus optional related sections (subscriber, patient, careDiagnoses, careRequestItems, careRequestDrugs, transportLocations, careRequestCase). The careRequest sub-object requires name and a request/record type (requestType or recordType); the request type must map to a CareRequest RecordType configured on the org. Multiple records can be created in one call by adding more list items.
-
-**Call:** `POST /connect/health/care-services`   — status: `implemented`
-
-**Inputs:**
-
-- `careRequestsList` *(`Array`)* — Top-level array of CareAuthorizationRequestInputRepresentation items, one per care request to create. Each item carries a careRequest sub-object (name + requestType/recordType required, plus serviceType/serviceLevel/sourceSystemIdentifier and other fields as needed) and optional related sections (subscriber, patient, careDiagnoses, careRequestItems, careRequestDrugs, transportLocations, careRequestCase).
-
-**Output:** CareAuthorizationRequestResultRepresentation {code, message, isSuccess, result[]}. Each result carries id (new CareRequest id), created (true when freshly made), status, objectType, and relatedRecords. Read the new record id from result[].id.
-
-#### `copy-care-request`
-
-Create a copy of a care request. Map the sections read by read-care-request into a new careRequestsList item (careRequest sub-object plus related sections) and POST it through the create path — dropping server-owned identity fields (id and any auto-assigned keys) so a genuinely new CareRequest is made rather than an update. Keep the source's request/record type so the copy still maps to a valid CareRequest RecordType. This is the create half of the read-then-create copy; there is no dedicated copy endpoint.
-
-**Call:** `postCareServices`   — status: `implemented`
-
-**Inputs:**
-
-- `careRequestsList` *(`Array`)* — One CareAuthorizationRequestInputRepresentation built from the source record's sections (careRequest sub-object + related sections), with server-owned id fields removed so a new CareRequest is created.
-
-**Depends on:** `read-care-request`
-
-**Output:** CareAuthorizationRequestResultRepresentation; result[].id is the copy's new CareRequest id, result[].created is true.
-
-#### `update-care-request`
-
-Update an existing care request. The update body uses the same top-level careRequestsList array as create (PATCH does NOT introduce a separate wrapper) and is sent as a PATCH to the collection path /connect/health/care-services — the record id travels inside the careRequest sub-object of the list item, not in the URL. Include the CareRequest id on the list item so the API updates rather than creates. The list item must ALSO carry a record type (careRequest.recordType or careRequest.requestType), exactly as a create does — the update re-resolves the record type, so an item with only an id and changed fields is rejected ("Invalid record type in this care request."). The published input schema marks recordType optional, but the service re-resolves it at runtime on every PATCH, so in practice you must include it. Pass the record's existing record type alongside the id and the fields to change.
-
-**Call:** `patchCareServices`   — status: `implemented`
-
-**Inputs:**
-
-- `careRequestsList` *(`Array`)* — Top-level array of CareAuthorizationRequestInputRepresentation items, each carrying the target CareRequest id AND a record type (recordType or requestType) on the careRequest sub-object, plus the fields to change. The record type is required on update (not just create) because the update re-resolves it.
-
-**Output:** CareAuthorizationRequestResultRepresentation; result[].created is false on an update (the record already existed).
-
-### Verify operations
-
-#### `verify-create-care-request`
-
-Confirm the care request was created by reading it back with the Care Services GET. Pass the new CareRequest id (result[].id from create-care-request) as the path id with idName=Id, and request the sections you want to confirm (semicolon-delimited sectionNames). A record coming back with the expected sections confirms the create landed.
-
-**Call:** `getCareServices`
-
-**Inputs:**
-
-- `ids` *(`String`)* — The new CareRequest id (result[].id).
-- `idName` *(`string`)*
-- `sectionNames` *(`string`)*
-
-**Depends on:** `create-care-request`
-
-#### `verify-copy-care-request`
-
-Confirm the copy created a distinct new CareRequest by reading the copied id back and checking it differs from the source id read in read-care-request.
-
-**Call:** `getCareServices`
-
-**Inputs:**
-
-- `ids` *(`String`)* — The copy's new CareRequest id.
-- `idName` *(`string`)*
-- `sectionNames` *(`string`)*
-
-**Depends on:** `copy-care-request`
-
-#### `verify-update-care-request`
-
-Confirm the update landed by reading the record back and checking the changed fields reflect the new values. Uses the same Care Services GET with idName=Id and the sections covering the fields that were updated.
-
-**Call:** `getCareServices`
-
-**Inputs:**
-
-- `ids` *(`String`)* — The updated CareRequest id (the id sent on the update list item).
-- `idName` *(`string`)*
-- `sectionNames` *(`string`)*
-
-**Depends on:** `update-care-request`