#Agentforce297 discussing

Hi @Gowtham V, 

Great analysis - what you’re seeing is expected.

The unmetered (per-user) benefit in Agentforce applies to UI/runtime usage, but Agent Script preview mode is treated as a dev/debug execution, which still consumes flex credits.

So: 

• UI usage -> no credits 

• Preview mode -> credits consumed

It’s not well documented, but this behavior is known. I'd suggest minimizing preview runs and raising a support case - this is a valid gap, especially since preview is essential for development. 

If you find this helpful feel free to mark it as Best Answer !!!! 😊 

Thanks and Regards.

Hi Avi, this is a common issue with the Agentforce Trailhead challenge. Here are the most common causes and fixes: 

1. Sign up for the correct Developer Edition org - You need a specific Agentforce-enabled Developer Edition org, not just a standard DE org. Go to the challenge instructions and use the sign-up link provided in the Trailhead module (usually linked as "sign up for a free Developer Edition org with Agentforce"). Standard orgs do not have Agentforce Builder enabled. 

2. Confirm the org is linked in Trailhead - In Trailhead, go to your profile and check under "Hands-On Orgs". Make sure the Agentforce DE org is set as the active org for this challenge. If it shows a different org, click "Connect Org" and authenticate with the Agentforce DE org credentials. 

3. Verify Agentforce is set up in that org - Log into your DE org and go to Setup, search for "Agents". If the Agents menu is missing, your org may not have the Agentforce feature enabled. In that case, sign up for a new org using the specific challenge sign-up link. 

4. Allow a few minutes after connecting - Sometimes the Trailhead checker takes 2-3 minutes to validate. Try clicking "Check Challenge" again after waiting. 

5. Hard refresh the Trailhead page - Press Cmd+Shift+R (Mac) or Ctrl+Shift+R (Windows) to force reload, then click Check Challenge again. 

If none of these work, open a Trailhead Help case at

using the Trailhead support option. 

Mani G | Principal/Founder |

Great question, Chris! Yes, Agentforce can absolutely be used to support or augment Business Acceptance Testing (BAT) on minor releases, though it works differently from traditional automated test frameworks. 

Here is how Agentforce can help with BAT: 

1 

. Agentforce Testing Center - Salesforce provides a dedicated Testing Center within Agent Builder where you can define test scenarios using natural language. You can create test cases that simulate user conversations and validate that your agent responds correctly to expected inputs and edge cases. 

2. Prompt Templates for Regression Scenarios - You can use Prompt Builder to define structured prompt templates that represent your BAT scenarios. These can be versioned and re-run after each release to confirm agent behavior has not regressed. 

3. Einstein Copilot / Agent Actions Testing - If your agents invoke custom Apex actions or flows, you can combine standard Salesforce testing (Apex tests, Flow testing) with Agentforce conversation-level testing to get end-to-end coverage. 

4. Bot Preview and Evaluation - During releases, use the agent Preview panel in Agent Builder to manually walk through critical user journeys before promoting to production. 

Current Limitations to Keep in Mind: 

- 

 Agentforce Testing Center does not yet support fully automated CI/CD pipeline integration out of the box (as of Spring 25). 

- For complex multi-step BAT scenarios, you may still need to complement with manual testing or third-party tools. 

R 

ecommended Steps: 

1. Go to Setup, search for Agents, open your agent in Agent Builder. 

2. Click the Testing tab and create test cases for your key release scenarios. 

3. After each minor release deployment, re-run those test cases to validate behavior. 

This gives you a repeatable, documented BAT process that leverages the AI agent natively. 

Mani G | Principal/Founder |

Thank you Manimaran Gunasekaran (Keneland)

!This is so helpful-  

I think I am following- So I could use just the one agent- IF I have the "escalation" topic- Just adding instructions to check business hours? Please see attached reference to escalation and the flow- 

Or do I need to add an action before routing to check the hours- 

Great question, Laura! The Agentforce In-Person Meeting Assistant is primarily designed for post-meeting analysis and CRM sync, not real-time Q&A during a live meeting.

Current capabilities include: automated post-meeting summaries, identification of product/pricing mentions, and syncing structured notes to Salesforce CRM via Conversation Intelligence (formerly Einstein Conversation Insights).

For real-time product Q&A during face-to-face meetings, the recommended approach is to configure a custom Agentforce Sales Agent grounded in your product catalog (via Data Cloud or a connected Knowledge Base) and enable it on the Salesforce Mobile App. This way, a rep can query the agent mid-meeting to surface specs, pricing, or competitive details on demand.

Recommended setup: (1) Build a custom Agentforce Agent with product catalog actions grounded in Data Cloud. (2) Enable it on the Salesforce Mobile App for live queries. (3) Use the Conversation Intelligence post-meeting flow to capture and sync the discussion to the Opportunity record.

This gives you both real-time lookup and automated post-meeting CRM sync.

-- Mani G | Principal/Founder | https://Keneland.com

Hi @Amanda Meneghini

Yes, store Testing Center results in a custom object -> create a report -> build a dashboard with charts. 

This is the standard Salesforce-recommended way to visualize any data. 

If you find this helpful, feel free to mark it as best answer !!!! 😊 

Thanks

Hi , 

Are the following two settings turned ON?