{"id":11471,"date":"2026-09-14T09:01:58","date_gmt":"2026-09-14T09:01:58","guid":{"rendered":"https:\/\/www.examlabs.com\/certification\/?p=11471"},"modified":"2026-09-14T09:01:58","modified_gmt":"2026-09-14T09:01:58","slug":"salesforce-certified-agentforce-specialist-practice-test-questions-and-exam-dumps-part2-q21-40","status":"publish","type":"post","link":"https:\/\/www.examlabs.com\/certification\/salesforce-certified-agentforce-specialist-practice-test-questions-and-exam-dumps-part2-q21-40\/","title":{"rendered":"Salesforce Certified Agentforce Specialist Practice Test Questions and Exam Dumps Part2 Q21-40"},"content":{"rendered":"

View Full\u00a0Salesforce Certified Agentforce Specialist Exam Dumps<\/a>\u00a0and Practice Test Dumps.<\/b><\/h2>\n

 <\/p>\n

Question 21<\/b><\/h3>\n

Which feature in Salesforce Data Cloud allows unstructured data like PDFs and knowledge articles to be indexed for Agentforce reasoning?<\/b><\/p>\n

    \n
  1. Vector Database and Hybrid Search<\/span><\/li>\n
  2. Standard Report Types<\/span><\/li>\n
  3. Schema Builder Relationship Trees<\/span><\/li>\n
  4. Standard Data Import Wizard<\/span><\/li>\n<\/ol>\n

    Correct Answer: 1<\/b><\/p>\n

    Explanation<\/b><\/p>\n

    Salesforce Data Cloud uses a built-in Vector Database alongside Hybrid Search capabilities to process and index unstructured enterprise data such as PDFs, support cases, and internal documentation. This technology converts raw textual content into numerical vector embeddings that can be semantically searched in real time. When an Agentforce agent receives a complex prompt, the Atlas Reasoning Engine queries this vector database to retrieve the most contextually relevant information chunks, allowing the agent to provide highly accurate, grounded responses without relying on manual data structuring or predefined database schemas.<\/span><\/p>\n

    Question 22<\/b><\/h3>\n

    What is the function of the “System Overview” instruction within an Agentforce Topic configuration?<\/b><\/p>\n

      \n
    1. It defines the database query timeout limits for Apex code execution<\/span><\/li>\n
    2. It provides high-level context to the LLM about when and why the topic should be selected<\/span><\/li>\n
    3. It sets the visual color scheme for the chat interface widget<\/span><\/li>\n
    4. It automatically generates user permission sets for the topic<\/span><\/li>\n<\/ol>\n

      Correct Answer: 2<\/b><\/p>\n

      Explanation<\/b><\/p>\n

      The System Overview instruction within an Agentforce Topic acts as the primary guiding prompt that explains the specific scope, boundary, and objective of that topic to the Large Language Model. When the Atlas Reasoning Engine processes an incoming user request, it evaluates this instruction to determine whether the user’s intent matches the topic’s domain. Writing a clear and comprehensive System Overview ensures that the agent accurately identifies when to route a conversation to this topic, preventing misclassification and ensuring the agent executes the correct actions assigned under that operational category.<\/span><\/p>\n

      Question 23<\/b><\/h3>\n

      Which type of Action can be added to an Agentforce Topic to collect structured data from a user via a step-by-step guided flow?<\/b><\/p>\n

        \n
      1. Screen Flow<\/span><\/li>\n
      2. Autolaunched Flow<\/span><\/li>\n
      3. Developer Console Query<\/span><\/li>\n
      4. Mass Email Action<\/span><\/li>\n<\/ol>\n

        Correct Answer: 2<\/b><\/p>\n

        Explanation<\/b><\/p>\n

        Autolaunched Flows are widely utilized as custom actions within Agentforce Topics to execute background business logic and step-by-step data processing. While Screen Flows require a direct human user interface to interact with, Autolaunched Flows allow the Atlas Reasoning Engine to programmatically collect parameters from the conversational transcript, pass those variables into the flow, execute complex backend logic, update CRM records, and return the execution results directly back to the agent. This seamlessly connects conversational user interactions with automated Salesforce platform business processes without requiring direct manual user input.<\/span><\/p>\n

        Question 24<\/b><\/h3>\n

        How does the Einstein Trust Layer handle sensitive information to prevent unauthorized data exposure to third-party LLM providers?<\/b><\/p>\n

          \n
        1. By storing sensitive data in external third-party servers<\/span><\/li>\n
        2. By utilizing dynamic Data Masking to replace PII with anonymized tokens before sending the prompt<\/span><\/li>\n
        3. By deleting all customer records involved in the prompt automatically<\/span><\/li>\n
        4. By encrypting the entire Salesforce organization during execution<\/span><\/li>\n<\/ol>\n

          Correct Answer: 2<\/b><\/p>\n

          Explanation<\/b><\/p>\n

          The Einstein Trust Layer protects enterprise privacy through dynamic Data Masking technology before any prompt is transmitted to external Large Language Model providers. It automatically scans the prompt payload for Personally Identifiable Information, such as social security numbers, credit card details, phone numbers, and email addresses, replacing them with anonymized placeholders. Once the LLM generates a response based on the masked data structure, the Trust Layer securely reverses the placeholders back into their original values before presenting the final response to the user, ensuring sensitive data never leaves Salesforce borders.<\/span><\/p>\n

          Question 25<\/b><\/h3>\n

          What is the purpose of testing an Agentforce Agent using the Agent Builder Inspector?<\/b><\/p>\n

            \n
          1. To compile Apex classes into production packages automatically<\/span><\/li>\n
          2. To simulate user conversations and trace how the Atlas Reasoning Engine evaluates topics, guardrails, and actions<\/span><\/li>\n
          3. To calculate monthly license subscription costs for Salesforce users<\/span><\/li>\n
          4. To automatically generate visual ERD diagrams for custom objects<\/span><\/li>\n<\/ol>\n

            Correct Answer: 2<\/b><\/p>\n

            Explanation<\/b><\/p>\n

            The Agent Builder Inspector provides administrators and developers with a real-time testing playground to validate agent behavior before deployment. It allows creators to input sample user prompts and observe a detailed step-by-step trace of the Atlas Reasoning Engine’s decision-making process. The inspector reveals which topic was triggered, how grounding context was retrieved, which specific actions were invoked, and how system guardrails shaped the final output. This visibility enables administrators to fine-tune topic instructions, adjust action parameters, and fix potential execution errors effectively.<\/span><\/p>\n

            Question 26<\/b><\/h3>\n

            Which capability allows Agentforce to generate personalized responses using dynamic Salesforce record data inside a Prompt Template?<\/b><\/p>\n

              \n
            1. Merge Fields and Data Resources<\/span><\/li>\n
            2. Static Text Strings<\/span><\/li>\n
            3. Hardcoded Apex Constants<\/span><\/li>\n
            4. HTML Style Sheets<\/span><\/li>\n<\/ol>\n

              Correct Answer: 1<\/b><\/p>\n

              Explanation<\/b><\/p>\n

              Prompt Templates in Salesforce leverage Merge Fields and dynamic Data Resources to ground Large Language Model outputs in real-time CRM data. By referencing specific field values, related records, or Data Cloud object attributes directly within the prompt construction, administrators can instruct the model to personalize its response based on the active user context. When executed by an Agentforce action, the engine automatically resolves these merge fields into actual record data, ensuring the generated text is precise, contextually tailored, and reflective of current system records.<\/span><\/p>\n

              Question 27<\/b><\/h3>\n

              What distinction exists between an Agentforce Action built using a Prompt Template versus one built using Apex?<\/b><\/p>\n

                \n
              1. Prompt Templates use generative AI for unstructured content creation, while Apex handles programmatic logic and procedural operations<\/span><\/li>\n
              2. Prompt Templates can only run on weekends, while Apex runs continuously<\/span><\/li>\n
              3. Apex cannot interact with standard Salesforce objects<\/span><\/li>\n
              4. Prompt Templates do not require any Salesforce setup permissions<\/span><\/li>\n<\/ol>\n

                Correct Answer: 1<\/b><\/p>\n

                Explanation<\/b><\/p>\n

                Prompt Templates and Apex serve complementary roles as actions within Agentforce. Prompt Templates utilize generative AI models to compose natural language text, summarize complex record histories, or draft context-aware communications using dynamic data inputs. Conversely, Apex actions execute rigid, deterministic, procedural code routines, such as performing advanced mathematical calculations, executing bulk transactional updates, or integrating with complex external legacy APIs. Combining both action types enables an Agentforce agent to handle both creative communication tasks and strict backend business process requirements smoothly.<\/span><\/p>\n

                Question 28<\/b><\/h3>\n

                Why is defining explicit “Guardrails” critical during the configuration of an Agentforce Agent?<\/b><\/p>\n

                  \n
                1. They increase the database storage allocation for custom objects<\/span><\/li>\n
                2. They establish strict operational boundaries to prevent the agent from providing unauthorized advice or off-topic responses<\/span><\/li>\n
                3. They convert all incoming text messages into standard PDF files<\/span><\/li>\n
                4. They bypass user security settings to speed up query execution<\/span><\/li>\n<\/ol>\n

                  Correct Answer: 2<\/b><\/p>\n

                  Explanation<\/b><\/p>\n

                  Guardrails establish essential safety boundaries and behavioural guidelines for an Agentforce Agent, ensuring it operates strictly within its designated organizational scope. By explicitly defining what the agent must not discuss\u2014such as providing financial advice, sharing internal corporate strategies, or using inappropriate language\u2014administrators prevent the AI from generating unauthorized, misleading, or harmful outputs. Combined with the Einstein Trust Layer, guardrails protect the enterprise brand image, maintain regulatory compliance, and ensure that conversational interactions remain professional, focused, and accurate.<\/span><\/p>\n

                  Question 29<\/b><\/h3>\n

                  What happens if an Agentforce Action fails to execute due to a system error during a live user chat?<\/b><\/p>\n

                    \n
                  1. The entire Salesforce instance shuts down immediately<\/span><\/li>\n
                  2. The agent handles the exception gracefully using fallback instructions or attempts alternative resolution paths<\/span><\/li>\n
                  3. The user’s account is permanently deleted from the database<\/span><\/li>\n
                  4. The system sends an email to all registered Salesforce customers<\/span><\/li>\n<\/ol>\n

                    Correct Answer: 2<\/b><\/p>\n

                    Explanation<\/b><\/p>\n

                    When an action failure occurs during a live conversation\u2014such as an API timeout or flow exception\u2014the Atlas Reasoning Engine is designed to handle errors gracefully. Rather than breaking the conversation session or exposing raw technical error messages to the user, the engine evaluates configured fallback protocols. It can inform the user of the temporary operational issue, ask alternative clarifying questions, retry the operation, or initiate a seamless transfer to a human support agent while logging the system error event for administrative review.<\/span><\/p>\n

                    Question 30<\/b><\/h3>\n

                    Which Salesforce tool is used to build custom, reusable Prompt Templates that can be invoked as actions by Agentforce?<\/b><\/p>\n

                      \n
                    1. Schema Builder<\/span><\/li>\n
                    2. Prompt Builder<\/span><\/li>\n
                    3. Process Builder<\/span><\/li>\n
                    4. Data Loader<\/span><\/li>\n<\/ol>\n

                      Correct Answer: 2<\/b><\/p>\n

                      Explanation<\/b><\/p>\n

                      Prompt Builder is the dedicated workspace within Salesforce where administrators and developers design, test, and manage custom Prompt Templates for AI features. It allows users to combine static natural language instructions with dynamic CRM merge fields, flow outputs, and Data Cloud object parameters. Once created and validated within Prompt Builder, these templates can be registered as discrete actions within Agentforce Topics, allowing agents to dynamically generate targeted communication, summaries, and complex text responses during user sessions.<\/span><\/p>\n

                      Question 31<\/b><\/h3>\n

                      How does Agentforce determine which Action to execute within a specific Topic during a conversation?<\/b><\/p>\n

                        \n
                      1. It executes every assigned action in alphabetical order automatically<\/span><\/li>\n
                      2. The Atlas Reasoning Engine analyzes the user’s intent and action descriptions to dynamically select the appropriate tool<\/span><\/li>\n
                      3. It randomly selects an action based on server availability<\/span><\/li>\n
                      4. The user must manually select the action name from a dropdown menu<\/span><\/li>\n<\/ol>\n

                        Correct Answer: 2<\/b><\/p>\n

                        Explanation<\/b><\/p>\n

                        The Atlas Reasoning Engine dynamically selects actions by evaluating the user’s explicit request against the natural language descriptions configured for each action within an active topic. Rather than following rigid, predefined decision trees, the engine interprets what the user is attempting to accomplish. If an action’s description matches the steps required to resolve the user’s prompt, Atlas automatically invokes that specific tool\u2014passing necessary context variables into it\u2014and uses the outcome to continue the conversational resolution process.<\/span><\/p>\n

                        Question 32<\/b><\/h3>\n

                        What is the role of the Zero-Data Retention policy in the Einstein Trust Layer?<\/b><\/p>\n

                          \n
                        1. It deletes customer records from Salesforce every 30 days automatically<\/span><\/li>\n
                        2. It guarantees that third-party LLM vendors do not store or retain enterprise prompt data after generating a response<\/span><\/li>\n
                        3. It prevents Salesforce administrators from viewing system debug logs<\/span><\/li>\n
                        4. It resets custom field values to null after every agent interaction<\/span><\/li>\n<\/ol>\n

                          Correct Answer: 2<\/b><\/p>\n

                          Explanation<\/b><\/p>\n

                          The Zero-Data Retention policy is a foundational privacy commitment embedded within the Einstein Trust Layer. When Salesforce transmits user prompts to external Large Language Model partners, contractual and technical safeguards ensure that these third-party providers do not store, log, or use enterprise customer data to train their commercial AI models. Once the model processes the prompt and returns the generated text, the data is immediately purged from the external provider’s memory, maintaining complete enterprise data governance.<\/span><\/p>\n

                          Question 33<\/b><\/h3>\n

                          An organization needs an Agentforce Agent to update customer billing addresses directly in CRM. Which component performs the actual database update?<\/b><\/p>\n

                            \n
                          1. The chat UI CSS script<\/span><\/li>\n
                          2. An Action (such as an Autolaunched Flow or Apex class) invoked by the agent<\/span><\/li>\n
                          3. The raw user prompt string<\/span><\/li>\n
                          4. The Vector Database indexer<\/span><\/li>\n<\/ol>\n

                            Correct Answer: 2<\/b><\/p>\n

                            Explanation<\/b><\/p>\n

                            While the Atlas Reasoning Engine interprets the user’s intent to update an address, it cannot alter database records directly without an execution tool. The actual DML operations are performed by an assigned Action\u2014such as an Autolaunched Flow or an Apex class exposed via <\/span>@InvocableMethod<\/span>. The reasoning engine passes the extracted new address details into the action parameters, which then executes standard Salesforce transaction logic to update the target CRM billing fields securely.<\/span><\/p>\n

                            Question 34<\/b><\/h3>\n

                            How can an administrator ensure an Agentforce Agent communicates using the organization’s specific brand tone and style?<\/b><\/p>\n

                              \n
                            1. By setting global agent instructions and persona definitions in Agent Builder<\/span><\/li>\n
                            2. By modifying the underlying Salesforce database code base<\/span><\/li>\n
                            3. By purchasing additional domain names for the Salesforce org<\/span><\/li>\n
                            4. By disabling generative AI capabilities across all topics<\/span><\/li>\n<\/ol>\n

                              Correct Answer: 1<\/b><\/p>\n

                              Explanation<\/b><\/p>\n

                              Brand alignment is established by configuring the agent’s Role, Persona, and general instructions inside Agent Builder. Administrators can explicitly define conversational parameters\u2014such as specifying whether the agent should sound formal, empathetic, or concise\u2014and instruct it on formatting preferences like bulleted steps or plain text. The Atlas Reasoning Engine incorporates these global instructions into every prompt context, ensuring all generated responses consistently reflect the enterprise’s official voice across all customer and employee touchpoints.<\/span><\/p>\n

                              Question 35<\/b><\/h3>\n

                              Which component allows Agentforce to perform Retrieval-Augmented Generation (RAG) using real-time corporate knowledge base articles?<\/b><\/p>\n

                                \n
                              1. Standard Workflow Rules<\/span><\/li>\n
                              2. Data Cloud Vector Search paired with Prompt Templates<\/span><\/li>\n
                              3. Developer Console Debugger<\/span><\/li>\n
                              4. Classic Knowledge Export Utility<\/span><\/li>\n<\/ol>\n

                                Correct Answer: 2<\/b><\/p>\n

                                Explanation<\/b><\/p>\n

                                Retrieval-Augmented Generation (RAG) is implemented in Agentforce by combining Data Cloud Vector Search with Prompt Templates. Knowledge base articles are indexed into vector embeddings within Data Cloud. When a user submits a question, the vector search retrieves the most relevant knowledge article fragments based on semantic similarity. These real-time information snippets are then injected into the prompt template via grounding resources, enabling the Large Language Model to compose accurate, contextually relevant answers grounded in enterprise documentation.<\/span><\/p>\n

                                Question 36<\/b><\/h3>\n

                                What security measure ensures that an Agentforce Agent does not display confidential financial records to an unauthorized user?<\/b><\/p>\n

                                  \n
                                1. The agent automatically redacts all numerical values in chat windows<\/span><\/li>\n
                                2. Salesforce Object-Level Security (OLS), Field-Level Security (FLS), and Sharing Rules enforced at runtime<\/span><\/li>\n
                                3. The third-party LLM vendor’s internal firewall rules<\/span><\/li>\n
                                4. Restricting agent chat access exclusively to administrative desktop computers<\/span><\/li>\n<\/ol>\n

                                  Correct Answer: 2<\/b><\/p>\n

                                  Explanation<\/b><\/p>\n

                                  Agentforce strictly operates within the native Salesforce platform security architecture. When an agent queries data or executes actions on behalf of a user, it honors the user’s specific Object-Level Security (OLS), Field-Level Security (FLS), and record-level Sharing Rules. If a user lacks read permissions for confidential financial records, the agent cannot access or ground responses in that data, preventing unauthorized data exposure and ensuring enterprise permission policies remain fully intact.<\/span><\/p>\n

                                  Question 37<\/b><\/h3>\n

                                  In Agentforce, what is the primary benefit of organizing capabilities into multiple distinct “Topics” rather than a single large topic?<\/b><\/p>\n

                                    \n
                                  1. It lowers the total cost of Salesforce user licensing<\/span><\/li>\n
                                  2. It improves reasoning accuracy by providing focused contexts, clear instructions, and relevant actions for specific user intents<\/span><\/li>\n
                                  3. It increases the file upload size limit for customer attachments<\/span><\/li>\n
                                  4. It eliminates the need for creating custom Salesforce fields<\/span><\/li>\n<\/ol>\n

                                    Correct Answer: 2<\/b><\/p>\n

                                    Explanation<\/b><\/p>\n

                                    Structuring an agent with multiple modular Topics drastically improves the precision of the Atlas Reasoning Engine. Categorizing capabilities into specialized domains\u2014such as “Billing”, “Technical Support”, or “Account Updates”\u2014allows administrators to provide clear, targeted instructions and restrict available actions to relevant operational boundaries. This separation prevents context confusion, reduces prompt execution complexity, ensures faster reasoning evaluation, and helps the agent select the exact tools required to fulfill specific customer requests correctly.<\/span><\/p>\n

                                    Question 38<\/b><\/h3>\n

                                    What role does the Audit Trail within the Einstein Trust Layer play in enterprise AI management?<\/b><\/p>\n

                                      \n
                                    1. It tracks developer changes made to custom Apex classes<\/span><\/li>\n
                                    2. It logs all AI interactions, masked data transformations, and safety evaluations for compliance and monitoring purposes<\/span><\/li>\n
                                    3. It automatically calculates employee payroll deductions based on chat duration<\/span><\/li>\n
                                    4. It archives expired user licenses to external storage drives<\/span><\/li>\n<\/ol>\n

                                      Correct Answer: 2<\/b><\/p>\n

                                      Explanation<\/b><\/p>\n

                                      The Einstein Trust Layer Audit Trail provides complete transparency into enterprise AI usage by capturing structured diagnostic logs for every generated response. It records essential metadata, including prompt execution timestamps, masked PII elements, toxicity detection scores, and LLM responses. Administrators and compliance officers can analyze these logs to monitor system usage, verify security policy adherence, evaluate safety guardrails, and audit generative AI decisions to ensure corporate governance and regulatory standards are met.<\/span><\/p>\n

                                      Question 39<\/b><\/h3>\n

                                      How does an Agentforce Agent initiate a handoff to a human service representative when a complex issue cannot be resolved automatically?<\/b><\/p>\n

                                        \n
                                      1. By terminating the user session and closing the browser window<\/span><\/li>\n
                                      2. By invoking a routing action that transfers the chat transcript to an Omni-Channel queue<\/span><\/li>\n
                                      3. By generating a random temporary password for the customer<\/span><\/li>\n
                                      4. By sending an unformatted text file to the server administrator<\/span><\/li>\n<\/ol>\n

                                        Correct Answer: 2<\/b><\/p>\n

                                        Explanation<\/b><\/p>\n

                                        When an agent encounters a request beyond its configured capabilities or detects explicit escalation criteria, it executes an escalation action integrated with Salesforce Omni-Channel. This action packages the active conversation history, summary context, and user details, routing the interaction to an available human agent’s queue. The service agent receives the complete conversational context within the Service Console, allowing them to resume support smoothly without forcing the customer to repeat information.<\/span><\/p>\n

                                        Question 40<\/b><\/h3>\n

                                        What is the recommended approach for validating an Agentforce Agent before deploying it to production environments?<\/b><\/p>\n

                                          \n
                                        1. Deploy directly to production and test live with active customers<\/span><\/li>\n
                                        2. Test extensively in a Sandbox environment using Agent Builder Inspector and user acceptance testing scenarios<\/span><\/li>\n
                                        3. Write standard workflow rules to replace all agent instructions<\/span><\/li>\n
                                        4. Disable all security settings temporary to perform rapid manual checks<\/span><\/li>\n<\/ol>\n

                                          Correct Answer: 2<\/b><\/p>\n

                                          Explanation<\/b><\/p>\n

                                          Before releasing an Agentforce agent to production, thorough testing must be conducted within a Sandbox environment. Administrators should utilize the Agent Builder Inspector to trace reasoning paths, verify topic selection logic, and validate action execution under various scenario inputs. Additionally, conducting User Acceptance Testing (UAT) with representative user personas ensures that security permissions, brand tone, fallback mechanisms, and human escalation handoffs function correctly without risking live customer interactions or production CRM data integrity.<\/span><\/p>\n","protected":false},"excerpt":{"rendered":"

                                          View Full\u00a0Salesforce Certified Agentforce Specialist Exam Dumps\u00a0and Practice Test Dumps.   Question 21 Which feature in Salesforce Data Cloud allows unstructured data like PDFs and knowledge articles to be indexed for Agentforce reasoning? Vector Database and Hybrid Search Standard Report Types Schema Builder Relationship Trees Standard Data Import Wizard Correct Answer: 1 Explanation Salesforce Data […]<\/p>\n","protected":false},"author":1,"featured_media":0,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":[],"categories":[1648,1647],"tags":[],"_links":{"self":[{"href":"https:\/\/www.examlabs.com\/certification\/wp-json\/wp\/v2\/posts\/11471"}],"collection":[{"href":"https:\/\/www.examlabs.com\/certification\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.examlabs.com\/certification\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.examlabs.com\/certification\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/www.examlabs.com\/certification\/wp-json\/wp\/v2\/comments?post=11471"}],"version-history":[{"count":1,"href":"https:\/\/www.examlabs.com\/certification\/wp-json\/wp\/v2\/posts\/11471\/revisions"}],"predecessor-version":[{"id":11472,"href":"https:\/\/www.examlabs.com\/certification\/wp-json\/wp\/v2\/posts\/11471\/revisions\/11472"}],"wp:attachment":[{"href":"https:\/\/www.examlabs.com\/certification\/wp-json\/wp\/v2\/media?parent=11471"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.examlabs.com\/certification\/wp-json\/wp\/v2\/categories?post=11471"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.examlabs.com\/certification\/wp-json\/wp\/v2\/tags?post=11471"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}