Updates to resolve open issues (#215)

* Fixes #165 updated images for SharePoint site creation

* Fixes #171 Web Search defaults to off, but is activated when adding URL knowledge items, students now will disable this option before testing their agent

* Fixes #173 - add step to Select Add channel

* fixes #175 - further changes to UI, only General Quality is there by default, added steps to configure an additional option

* fixes #176 - increased to 3-5 minutes

* fixes #177 - updated to reflect ui

* removed image for now

* Fixes #182 - fixed typo

* fixes #185 - changed chit chat example

* Fixes #188 - fixed typo Topic.Formula should have been Topic.relevantNewsForOpportunities

* Fixes #189 - removed reference to non old steps

* fixes #200 - fixed names to be consistent

* fixes #201 - added note

* fixes #184 #183 - simplified commission calculation

Author: davidyack

labs/autonomous-account-news/README.md

@@ -231,6 +231,7 @@ Set up an autonomous agent with a recurring trigger that automatically activates
         ```
         cat_amount gt 300000 and cat_isclosed eq false
         ```
+    You may see a validation error like missing required property - value. After a few seconds or as you select away from that input field it should clear. Also, it is not necessary to select the ... to provide the OData filter.
 
   1. Select **Save** to finalize the tool configuration.
 
@@ -417,7 +418,7 @@ Set up an autonomous agent with a recurring trigger that automatically activates
       - Select on **(+)** to add a node. 
       - Select **Variable management** and then **Set a variable value**.
       - For **Set variable**, create a new variable, make it **Global**, and name it `relevantNewsForOpportunities`. 
-      - In **To value** select `Topic.relevantNewsForOpportunities`. You need to select ... then select Formula, then type **Topic.Formula**, then select Insert.
+      - In **To value** select `Topic.relevantNewsForOpportunities`. You need to select ... then select Formula, then type **Topic.relevantNewsForOpportunities**, then select Insert.
 
   1. Select **Save**
 
@@ -481,9 +482,6 @@ Set up an autonomous agent with a recurring trigger that automatically activates
   5. Use Log relevant news for opportunities to log your findings. The base input for Log relevant news for opportunities should be {Global.searchResults} with determined relevance added
   ```
 
-> [!TIP]
-> If instructions don't offer the "/" option to reference topics, tools or variables, you can skip step 26-30 and continue with the next steps. You can always come back to this later.
-
   1 To increase orchestration accuracy, you will now replace names of topics, tools and variables with references. References can be added to instructions by typing **/** and selecting the appropriate object from the drop-down menu.
 
   1. In the instructions, select `<Get Opportunity records>`. Type **/** and in the drop-down menu, under **Tool**, select **Get Opportunity records**. The previous text in curly brackets should be replaced by a visual reference to the tool.

labs/core-concepts-agent-knowledge-tools/README.md

@@ -263,6 +263,12 @@ Add a document knowledge source to your agent and verify that it accurately answ
 
 1. Review the **Name** and **Description** fields. Update if needed to make the source easily identifiable.
 
+#### Check and disable Web Search 
+The Use information from the web setting is available on the Generative AI settings page or the Web Search setting in the Knowledge section of the agent's Overview page. This setting lets your agent access broad, real-time, and up-to-date information beyond what is available in predefined or enterprise-specific knowledge bases. For our scenario, we want to keep the use of knowedge focused on our provided resources and not the broader web.
+
+1. Navigate to the Overview tab, scroll down to the Knowledge section
+
+1. Select Disabled on the Web Search option.
 
 #### Test Knowledge Integration
 

labs/core-concepts-analytics-evaluations/README.md

@@ -319,6 +319,14 @@ Create evaluation test sets using four different methods and understand how each
 
 1. In the **Configure test set** panel on the right side of your screen, change the test set name to **Non-Critical Copilot Studio Guide Set**
 
+1. In the Test method General Question is configured by default. You can configure additional test methods, those will be available to configure later when you edit questions.
+
+1. Select **Add test method**, review the list of test methods you can configure.
+
+1. On the **Set pass score**, review the settings for this option and select **OK**. Other test methods may have other options you can configure based on the approach.
+
+1. Select **Compare meaning** and then select **OK**.
+
 1. Select **Save** at the bottom of that same panel.
 
 1. In that same panel, select the **Manage profile** button.
@@ -329,25 +337,15 @@ Create evaluation test sets using four different methods and understand how each
 
 1. Select the first generated question to explore all the available options. For each test case, you can configure the **evaluation method**:
 
-    - **Exact Match**: Character-for-character comparison between expected and actual response. Use for questions with precise, factual answers.
-
-    - **Keyword Match**: Checks whether key terms from the expected response appear in the actual response. Use when exact wording doesn't matter but key concepts must be present.
-
-    - **Similarity**: Uses cosine similarity to compare semantic meaning on a 0-1 scale with a configurable threshold. Use when meaning matters more than exact wording.
-
-    - **General Quality**: Uses a LLM to evaluate response quality across four dimensions - relevance, groundedness, completeness, and abstention. Does NOT require an expected response. Use for open-ended questions.
-
-    - **Compare Meaning**: Evaluates whether the intent and meaning of the actual response matches the expected response, with a configurable threshold. Use for semantic comparison with more nuance than cosine similarity.
-
     > [!TIP]
-    > Choose evaluation methods that match the nature of each question. Factual questions with precise answers work well with Exact Match or Keyword Match. Open-ended questions benefit from General Quality or Similarity methods.
+    > Choose evaluation methods that match the nature of each question. Factual questions with precise answers work well with Exact Match or Keyword Match. Open-ended questions benefit from General Quality or Similarity methods. You can add additional test methods to the set at creation of the test set or later during editing.
 
 1. After reviewing, select **Cancel** to close the edit of the test case.
 
 1. Select **Evaluate** to start the evaluation of this test set.
 
     > [!NOTE]
-    > Evaluation time depends on the number of test cases and agent response time. A test set with 10 cases typically completes in 1-3 minutes.
+    > Evaluation time depends on the number of test cases and agent response time. A test set with 10 cases typically completes in 3-5 minutes.
 
 1. After your evaluation runs, review the overall result and then select the evaluation row to drill down into details for each question.
 
@@ -520,10 +518,8 @@ Review and interpret evaluation results, compare outcomes across test sets, and
 
 1. Select an individual test case to view its detailed results:
    - **Question**: The original test question
-   - **Expected response**: What the AI generated as the correct answer
    - **Actual response**: What the agent actually responded with
    - **Result**: Pass or fail
-   - **Reasoning**: An explanation of why the test passed or failed
 
 1. For any failed test cases, review the **activity map** to see the step-by-step conversation flow showing the agent's decision path, including which knowledge sources, tools, and topics were used.
 

labs/core-concepts-variables-agents-channels/README.md

@@ -501,6 +501,8 @@ Deploy your agent to Teams and Microsoft 365 Copilot channels with proper config
     - **App icon**: Visual branding in Teams
     - **Availability**: Who can access the agent
 
+1. Select **Add channel** in the bottom right corner of the panel to start activation of this channel.
+
 1. Select **Edit details** and make adjustments. For example, change the short description to something like **Assist users building agents**
 
 1. Select **Save** to save your changes.

labs/mbr-prep-sharepoint-agent/README.md

@@ -160,7 +160,7 @@ Create a SharePoint site with MBR documents that will serve as the knowledge sou
 
 ![Create Site](./assets/site-name.png)
 
-7. Click **Finish**
+7. Select **Create site**
 
 ![Select Finish](./assets/select-finish-site-creation.jpeg)
 

labs/mbr-prep-sharepoint-agent/assets/select-finish-site-creation.jpeg

@@ -378,8 +378,6 @@ Create a deployment pipeline that automates solution deployment across environme
 
 1. **Save** the pipeline configuration.
 
-    ![Pipeline configuration](images/pipeline.png)
-
 #### Test Your Pipeline
 
 1. In the PROD card, select **Deploy here**.

labs/mbr-prep-sharepoint-agent/assets/site-name.png

@@ -529,7 +529,7 @@ Connect existing agent to our Sales Associate Assistant agent to add ability to
 
 #### Connect the Account and Contact Information Agent
 
-1. In the Copilot Studio tab in your browser, go to the parent agent (Sales Info Assistant)
+1. In the Copilot Studio tab in your browser, go to the parent agent (Sales Associate Assistant)
 
 1. Go to the **Agents** menu on the top navigation bar