Skip to main content
Gruppo

Agentblazer Community Group

Welcome, Agentblazer! Are you ready to connect with AI leaders and enthusiasts from around the world? You've come to the right place. The Agentblazer Community is the premier destination to skill up on AI, discuss best practices, gain access to Agentforce product experts, showcase your expertise, and supercharge your career.

Episode 1 Challenge

 

If someone ran Org Check on your Salesforce org right now without warning. 

what's the one thing you'd be quietly hoping they don't find? 

 

The field nobody knows the purpose of. 

The Apex class with no test coverage. 

The permission set that grants way too much. 

The Flow that runs daily and whose logs you've never checked. 

 

Mine: a Workflow Rule that's been active since 2011 on an object that nobody uses anymore. Still firing. Every save. For no reason. 

 

What's yours? 

 

#NoClicksJustVibes #DeFraggingTheFrankenOrg@Salesforce Developer Group, Dubai, UAE @WFD Partner Cohort: TDX2630DaysChallenge @Agentblazer Community Group @* Salesforce Developers *

4 commenti
0/9000

I have hundreds of customer surveys in my Salesforce (in a custom object Customer Survey) 

Is it a good idea to create an agent to analyse for exemple the 500 surveys of this quarter in a single shot ? 

My marketing team is asking for that. 

Or shoud I use another tool ?

4 risposte
  1. 3 giu, 10:47

    @Hugues Raguet I would not use Agentforce to analyze 500 survey records in one single shot.

    Agentforce is better for guided, contextual work: answering a user’s question, summarizing a specific customer/account, helping a rep take action, or analyzing a controlled set of records. For bulk survey analysis, you can quickly run into prompt size, performance, cost, governor-limit, and consistency issues. Salesforce Prompt Builder also has limits, including a maximum prompt template size of 128,000 characters, so pushing hundreds of survey responses into one prompt is not a stable design.  

    Check:  Salesforce Help

      

    A better architecture is to process the surveys in batches. For example, use Flow, Batch Apex, Data Cloud, Tableau, or CRM Analytics to prepare the dataset first. Then use AI to summarize smaller chunks, such as per Account, per product, per region, or per month. Store those summaries or sentiment/theme fields back on records, then let Agentforce answer questions from that prepared/grounded data. 

     

    Salesforce has a survey feedback summarization use case that follows this kind of pattern: summarize free-format survey feedback using Flow-grounded prompt data, rather than asking an agent to analyze every raw survey in one huge request.  

    Check: Analyse Survey Feedback | Salesforce AP

0/9000

🆕 Trailhead Quest: Agentic Workflow Automation — Now Live!

 

Ready to build with AI? The Agentic Workflow Automation Quest is live all month — and we want to see what you create. 

 

What you'll do:

  • Complete the Trailhead trailmix
  • Build an agentic workflow using Prompt Builder + Agentforce Grid
  • Share your use case via the Google Form

 

🏆 Enter for a chance to win* 1 of 65 Salesforce certification vouchers. 

🗓️ Deadline: May 31, 2026 

 

 👉 Start the Quest

 

*Restrictions apply.  

 

See Official Rules. http://sfdc.co/bP7smx🆕 Trailhead Quest: Agentic Workflow Automation — Now Live! Ready to build with AI? The Agentic Workflow Automation Quest is live all month — and we want to see what you create.

9 commenti
  1. Oggi, alle 06:58

    Hi @N R it's isn't like that a list will be coming, if u are one of them u will receive a mail , if not then u're not the winner .

0/9000

Hi Team I was practising https://trailhead.salesforce.com/content/learn/projects/customize-a-service-agent-with-prompts-flows-and-actions/update-a-flow-agent-action-and-topic?trail_id=become-an-agentblazer-innovator-2026

this trailhead and I saw get sessions method being called 

 

actions:

Get_Sessions: @actions.Get_Sessions

with experienceId = ...

with startDate = @variables.converted_date

        although in trailhead it was not mentioned to try but i wanted to see how to pass  vairables to actions .I observed that irrespective of variable value available or not it always passes null 

in other actions slot filling mechanism is used .  

Get_Experience_Details: @actions.Get_Experience_Details

with experienceName = ...

    As per observations it seems to call any actions we have to use slot filling only variables cannot be used.  

Community please correct me if i am doing wrong

5 risposte
  1. Oggi, alle 06:35

    As being said, the variable must be populated before

    you use it in action. To set a variable, use this notation: 

     

            actions:

                set_start_date: @utils.setVariables

                    with startDate = ...

     

    Instructions should tell the LLM what date is expected and its format. 

0/9000

RSVP : https://trailblazercommunitygroups.

 

RSVP :

https://trailblazercommunitygroups.com/events/details/salesforce-salesforce-developer-group-dubai-uae-presents-de-fragging-the-franken-org-session-2-the-attic-nobody-wanted-to-open/

 

 

Your Salesforce metadata is the DNA of your org. Custom fields, Apex classes, Flows, Validation Rules, Profiles — every piece of configuration tells a story. After 10 years? Some of those stories are concerning. 

 

In this session we go through the Metadata Lens — identifying what's deployed, what's dead weight, what's dangerous for Agentforce agent actions, and what needs to be cleaned or restructured before you build on top of it. 

 

What you will learn: 

 

  • The 6 metadata categories that matter most for Agentforce readiness
  • How to run Salesforce Code Analyzer for static code analysis — find severity-1 violations before they break your agent actions
  • The Safe Deletion Protocol — how to remove unused metadata without causing a 3am production incident
  • Why Unlocked Packages are the 2026 standard for metadata organization and how to create your first one
  • Apex code quality standards specific to Agent Actions — what must be fixed before your code becomes agent-callable
  • How to prioritize technical debt: fix now vs. fix before Agentforce vs. backlog
  • Tools Covered: Salesforce Code Analyzer (CLI plugin), OrgCheck , Salesforce CLI metadata commands, Unlocked Package creation

Who Should Attend: Salesforce Architects, Developers, Senior Admins managing legacy org technical debt 

Prerequisites: Session 1 recommended but not required. Basic understanding of Salesforce metadata helpful. 

 

@Salesforce Developer Group, Dubai, UAE @WFD Partner Cohort: TDX2630DaysChallenge @Agentblazer Community Group @* Salesforce Developers * @Architect Trailblazers #NoClicksJustVibes #DeFraggingTheFrankenOrg

0/9000
2 commenti
0/9000

🎙️ Episode 1 is on YouTube — De-Fragging the Franken Org 

 

"Your Legacy Org Has Feelings - And We're About to Find Out Why" 

 

Missed the live session? check it here : 

https://www.youtube.com/watch?v=OwmZhPJ1V9Q

 

→ What Agentforce-Ready means in 2026 (no buzzwords) 

→ The 8-Lens modernization framework explained 

→ Org Check + Salesforce Inspector Reloaded + Salesforce CLI — live run 

→ Org Complexity Scorecard built in real time 

→ Final score: 17/40 — and what it means 

 

What's your org's score? Drop it below. 👇 

 

#NoClicksJustVibes #NoClicksJustVibes#NoClicksJustVibes #DeFraggingTheFrankenOrg@Salesforce Developer Group, Dubai, UAE @WFD Partner Cohort: TDX2630DaysChallenge @Agentblazer Community Group @* Salesforce Developers * @Architect Trailblazers

0/9000

Whenever any inappropriate input or query comes to AI Agent, Salesforce Automatically routing to to "Inappropriate_Content" subagent and giving a standard message. Is there any way can we respond with our own custom message and terminate the session?

1 risposta
  1. 5 giu, 11:54

    Hi @Ravali Reddy,

    In your use case requires a custom response, you could use handling those scenarios through your own topics, instructions, or custom routing logic before the request reaches that point. Whether the standard Inappropriate_Content response can be completely overridden may depend on the specific Agentforce configuration. 

0/9000

Hi! Is it possible to customize Inappropriate Content

topic to have agent answer with a different message?  

 

Tried to add an instruction to a different topic for agent to request the customer to rephrase the question, but it is not working and still returns "Sorry, I can't assist with that. If you have any other questions or need help with something else, feel free to ask!" 

 

Thanks! 

 

 

2 risposte
0/9000

Is it feasible to create custom button inside the chat window of internal Agentforce (Copilot) which can call the Agent actions in the backend ,Instead of user having to type the prompt, can they just click one button?? Similar to "summaries opportunity" button shown in the below screenshot is available by default on opportunity.  

 

My requirement is for internal Agenforce which is the default agent(Copilot) and it is not exposed to salesforce sites or web channels. Internal salesforce users with the access would be clicking on the Agentforce icon near the setup and they will get the Agent force chat window.. i want to add custom button which calls custom Agent action inside that chat instead of typing the prompt.. 

 

Similar to below out of the box "summaries opportunity" clickable icon. Can we create custom button inside Agentforce Copilot chat window?

 

 

3 risposte
  1. 29 ago 2025, 13:36

    @Akshay Mense any luck with this? I'm also looking for a way to add custom buttons to the Employee Agent chat window. Here's an idea about this feature in general. At the same time, it seems we can now show custom LWCs in the chat.

    Page buttons might work if your question is in the context of the currently opened record. But the chat window might be accessed from anywhere in the org, so I'd like to have custom buttons directly in the chat. 

0/9000