arohitu/salesforce-revenue-cloud-skills

revenue-cloud-decision-table

Finds and verifies Salesforce Revenue Cloud Decision Tables.

ソースを見る

使い方

  1. インストールコマンドをコピーし、自分の環境へ導入する前にソースを確認します。
  2. 必要なツール、認証情報、対応環境を原文で確認します。
  3. 入力と成果物を指定し、出力を資料と照合します。
npx skills add https://github.com/arohitu/salesforce-revenue-cloud-skills --skill revenue-cloud-decision-table

開始前の確認

  • 導入にはモデル契約、API枠、ファイルへのアクセス許可は含まれません。
  • スター数や導入数は活動の指標であり、独立した品質・安全性試験ではありません。
リポジトリの原文

見出し、例、コード、表、リンク、参照画像を含む原文を表示しています。

Revenue Cloud Decision Table

Use this skill to find a Salesforce Revenue Cloud Decision Table, determine the inputs needed to invoke it, collect missing runtime values, and verify what the table returns from the org.

Default Workflow

  1. Identify the target org alias or username. If missing, ask for it or use the Salesforce CLI default org only after saying so.
  2. Find the Decision Table Id from the user's table name, developer name, label, pricing component, or business scenario.
  3. Inspect the table metadata and parameters to identify input fields, output fields, operators, and whether a dataset link name is required.
  4. Build a conditions payload. If runtime values are missing, ask the user for only the missing input values; otherwise infer them from provided quote/order/product/pricing context.
  5. Invoke the table through the Connect REST API and report the returned outcomeList[].values, outcomeType, successStatus, and any errors.
  6. Compare the outcome with the user's expected pricing behavior and call out mismatched inputs, wrong dataset link, inactive table, stale refresh status, or auth problems.

Helper Scripts

The scripts require Salesforce CLI sf, jq, and an authenticated target org.

  • scripts/find-decision-table.sh: finds candidate Decision Tables by search text or developer name.
  • scripts/inspect-decision-table-inputs.sh: queries DecisionTableParameter metadata and emits a payload skeleton with null values for inputs the agent must fill.
  • scripts/invoke-decision-table.sh: uses sf org display to get the org instance URL and access token, then posts a payload to the Decision Table invocation endpoint.

Run scripts from this skill folder or pass their full path:

bash
.cursor/skills/revenue-cloud-decision-table/scripts/find-decision-table.sh --target-org <alias> --search "Volume"
.cursor/skills/revenue-cloud-decision-table/scripts/inspect-decision-table-inputs.sh --target-org <alias> --decision-table-id 0lD...
.cursor/skills/revenue-cloud-decision-table/scripts/invoke-decision-table.sh --target-org <alias> --decision-table-id 0lD... --payload-file payload.json

Never write real access tokens to files or include them in final answers.

Finding The Decision Table Id

Use the script first when Salesforce CLI access is available:

bash
scripts/find-decision-table.sh --target-org <alias> --search "<name or label>"

If scripting is not available, use SOQL:

soql
SELECT Id, DeveloperName, MasterLabel, Type, Status, SourceObject,
       ExecutionType, UsageType, RefreshStatus
FROM DecisionTable
WHERE DeveloperName = '<DecisionTableDeveloperName>'

For fuzzy discovery, search DeveloperName and MasterLabel with LIKE. Confirm:

  • Id is the runtime id used in the URL.
  • Status is active for runtime pricing.
  • UsageType and Type match the pricing/business scenario.
  • RefreshStatus indicates data is usable.
  • SourceObject matches the expected backing table data.

Identifying Inputs

Use:

bash
scripts/inspect-decision-table-inputs.sh --target-org <alias> --decision-table-id <0lD...>

Read the returned suggestedPayload. Replace every null value with a runtime value. If the script cannot classify parameters cleanly, inspect parameters and ask the user for values for input-like fields. Use the source quote, order, product, price book, pricing scheme, quantity, tier, and attribute values when the user already provided them.

Request only the missing values, for example:

text
I found RCA_VolumeBasedPrices with these required inputs:
RCA_BundleProduct__c, RCA_ComponentProduct__c, RCA_PriceBook__c, RCA_PricingScheme__c, RCA_LowerBound__c, RCA_UpperBound__c.
Please provide values for RCA_ComponentProduct__c and RCA_PriceBook__c.

Invocation Endpoint

Use Decision Table Invocation unless the user specifically asks for the legacy lookup endpoint:

http
POST https://<instance>.salesforce.com/services/data/v66.0/connect/business-rules/decision-table/lookup/<decisionTableId>
Authorization: Bearer <access_token>
Content-Type: application/json

Use API version 58.0 or later; prefer the org's current supported version. The legacy Decision Table Lookup resource is:

http
POST /services/data/v55.0/connect/business-rules/decision-table/<decisionTableId>

Payload Pattern

json
{
  "datasetLinkName": "RCA_VolumeBasedPrices_Default",
  "conditions": [
    {
      "conditionsList": [
        {
          "fieldName": "RCA_BundleProduct__c",
          "value": "01tVc00000CojkbIAB",
          "operator": "Equals"
        },
        {
          "fieldName": "RCA_LowerBound__c",
          "value": 5,
          "operator": "GreaterOrEqual"
        },
        {
          "fieldName": "RCA_UpperBound__c",
          "value": 5,
          "operator": "LessThan"
        }
      ]
    }
  ]
}

Rules:

  • conditions is required.
  • datasetLinkName is optional, but include the dataset link API name when the table uses one. Common naming is <DecisionTableDeveloperName>_Default; verify where possible.
  • fieldName must match the Decision Table input field API name.
  • value must use the correct JSON type: numbers unquoted, booleans as booleans, strings quoted.
  • operator is optional and overrides the table's configured operator when supplied.
  • Add sourceObject when a dataset link maps multiple source objects and the field is ambiguous.

Common operators: Equals, NotEquals, GreaterThan, GreaterOrEqual, LessThan, LessOrEqual, Matches, ExistsIn, DoesNotExistIn.

Verify The Return

After invoking, inspect:

json
{
  "errorCode": null,
  "errorMessage": null,
  "outcomeList": [
    {
      "values": {
        "RCA_UnitPrice__c": "23043.04"
      }
    }
  ],
  "outcomeType": "Single Match",
  "successStatus": true
}

Some invocation responses wrap outcomes in outputs[]; handle both a direct Decision Table Outcome and an outputs array. Treat outcomeList[].values as the returned table outputs.

Report the verification as:

markdown
Decision Table: <DeveloperName> (`<Id>`)
Inputs tested: <field=value, ...>
Outcome type: <Single Match|Multiple Matches|...>
Returned values: <key outputs>
Status: <success or errorCode/errorMessage>

Troubleshooting

For INVALID_AUTH_HEADER:

  • Confirm the actual HTTP header is Authorization: Bearer <access_token>.
  • Replace placeholders such as {{your_access_token_here}}; do not include braces in the header value.
  • Ensure the token belongs to the same Salesforce instance used in the request URL.
  • Refresh expired tokens and confirm the user has API and Decision Table access.

For no match or wrong match:

  • Confirm the Decision Table is active and refreshed.
  • Verify datasetLinkName is an API name, not a label.
  • Check input field API names, operators, value types, and source object mapping.
  • For ranges, validate inclusive/exclusive operator pairs such as GreaterOrEqual with LessThan.
  • Compare the API payload to runtime pricing context values after hydration, Apex hooks, attribute flattening, and formula preparation.

Apex Alternative

Inside Salesforce Apex, use ConnectApi.DecisionTableInput, add ConnectApi.DecisionTableCondition values, optionally set input.datasetLinkName, and call:

apex
ConnectApi.DecisionTableOutcome output =
    ConnectApi.DecisionTable.execute(decisionTableId, input);

タスクに関連する項目

同じリポジトリから

関連する Skills

すべての Skills
arohitu
コミュニティ

revenue-cloud-config-apis

Reference and call the Salesforce Revenue Cloud (Revenue Lifecycle Management) Product Configurator Business APIs. Use when working with the CPQ configurator Connect REST resources - configuring a product bundle, loading/getting/setting/saving a configuration instance, adding/updating/deleting nodes, setting product quantity, or running configuration rules - and when the user mentions Revenue Cloud, RLM, Product Configurator, CPQ configurator, /connect/cpq/configurator, contextId, transactionId,

導入数
1
GitHub Stars
21
更新日
9月1日
arohitu
コミュニティ

revenue-cloud-pcm

Salesforce Revenue Cloud Product Catalog Management (PCM) master data on core Salesforce objects for Agentforce Revenue Management, Revenue Cloud Advanced, and Revenue Cloud Billing. Use for any lookup, query, design, build, export, import, migration, CSV, sandbox load, troubleshooting, or analysis involving catalogs, categories, products, bundles, classifications, dynamic attributes, picklists, qualification/disqualification, selling models, Product Discovery, Product2, ProductCatalog, ProductC

導入数
1
GitHub Stars
21
更新日
9月1日
arohitu
コミュニティ

revenue-cloud-pricing-diagnostics

Use this skill when dissecting Salesforce Revenue Cloud pricing in core Revenue Cloud orgs, tracing how a Quote, Order, Quote Line, or pricing field is populated, debugging pricing procedures, context definitions, context mappings, expression sets, pricing elements, decision tables, procedure plans, or Apex pricing hooks. Use it for Revenue Cloud pricing analysis even when the user asks generally how a price, total, discount, exchange rate, or field value is calculated. Do not use it for Salesfo

導入数
1
GitHub Stars
21
更新日
9月1日
arohitu
コミュニティ

revenue-cloud-txn-management-apis

Reference and call the Salesforce Revenue Cloud (Agentforce Revenue Management) Transaction Management Business APIs - the Connect REST Quote & Order Capture resources that place quotes and orders, amend/renew/cancel/swap/upgrade/downgrade assets, fetch instant pricing, clone or read a sales transaction, create supplemental/change orders, manage ramp deals, and get or create promotions. Use when the user wants to create or update a quote or order via REST, price a quote/order line data grid, run

導入数
1
GitHub Stars
21
更新日
9月1日