Yes thanks for the passionate response and agree that returning lots of records to agent wouldn't be great.
In my example that was not going to be the case. The LLM is doing a good job translating natural language prompts into a SOQL query with this action. That's the piece I want. The output of records would be done with Flow or Apex and does not need to go to the LLM other than maybe 1 variable with the total records.
Yes I can, through trial and error eventually do this myself but it seemed unnecessary since their action is doing that but doesn't look like I can see any instructions or .MD as to how SF is doing that so I'll need to reinvent the wheel on this.
50 is still pretty small for others though and if they want to use up tokens returning 100-200 I'd still vote for that option. 200 is usually the magic # in SF for records.
Agentblazer Community Group
Once you start building complex agents, you’ll soon run into this issue:
Your agents will not consistently ‘decide’ the right path. Example: instead of escalating to human agents it will keep responding . (Though you may have intent expressed in your instructions or tried bifurcating with sub-agents or actions)
So how do you solve this use case? We went live with an agent and this is what we learned how to structure it.
https://saisocial.substack.com/p/agentforce-how-to-structure-decision
@Buyan thyagarajand: thank you, the best teachers are actual client implementations. By the time it's in prod, you'll learn lessons no trailhead can teach.
🚀 Salesforce Winter ’27: TOP New Features for Developers & LWC!
Salesforce Winter ’27 is bringing some exciting updates for Salesforce Developers and Lightning Web Components (LWC). 🔥
In my latest video, I cover the TOP Winter ’27 features that developers should know, including:
💻 New developer capabilities
⚡ Lightning Web Components (LWC) improvements
🔌 API & platform enhancements
🚀 Performance and scalability updates
🔐 Security & governance improvements
🛠️ Features that can make development faster and more productive
If you're a Salesforce Developer, LWC Developer, or working with the Salesforce Platform, these updates are definitely worth checking out!
🎥 Watch the video:
https://www.youtube.com/watch?v=8j97AenYLpM
📺 Subscribe to Salesforce Hunt:
https://www.youtube.com/@SalesforceHunt
@Agentblazer Community Group @* Sales Cloud - Getting Started *@* Release Readiness Trailblazers *@* Sales Cloud - Best Practices *@* Salesforce Administrators *@* Service Cloud *
🚀 Salesforce Winter ’27: Exciting New Features for Flow Builders! ⚡
Salesforce Winter ’27 is bringing some great improvements to Flow — making automation easier to build, debug, maintain, and scale.
Here are some of the Flow features every Salesforce Admin & Developer should know 👇
🔹 1. Improved Flow Builder Experience
Build and manage your flows more efficiently with enhancements to the Flow Builder experience.
🔹 2. Better Flow Debugging & Troubleshooting
🐞
New improvements make it easier to understand what’s happening during flow execution and identify issues faster.
🔹 3. Enhanced Flow Performance
⚡
Salesforce continues to improve Flow capabilities so you can build more scalable automation while keeping performance in mind.
🔹 4. More Powerful Flow Logic
🧠
New enhancements provide additional flexibility when designing complex automation and business logic.
🔹 5. Improved Flow Management
Managing multiple versions and maintaining automation becomes easier with continued improvements across the Flow lifecycle.
💡 Why does this matter?
Flow is becoming more than just a point-and-click automation tool. With every release, Salesforce is making Flow a stronger platform for building enterprise-grade automation with less custom code.
👉 If you're working with Salesforce Flow, Winter ’27 is definitely a release worth exploring.
Which Winter ’27 Flow feature are you most excited about? 👇
🎥
Want to see these features in action?I’ve covered the Salesforce Winter ’27 Flow features in this video:
👉 https://www.youtube.com/watch?v=8j97AenYLpM&t=348s
@Agentblazer Community Group @* Sales Cloud - Getting Started *@* Release Readiness Trailblazers *@* Sales Cloud - Best Practices *@* Salesforce Administrators *@* Service Cloud *
"General," "General 2," "General Support," and "General (DO NOT DELETE)."
That's a real Subagent list. Well real-ish. Episode 3 of Season 3: The Agentforce Architects is live, and it's all about keeping Subagents, Actions, and Instructions clean as your Agentforce org actually scales.
Inside this episode:
- 🗂️ A naming convention that scales: Domain – Capability – Risk Tier
- 🔁 Why Actions should be reused, not rebuilt, four times with slightly different names
- 📝 Writing Instructions as testable conditional rules instead of vague vibes
- 🎬 The true story of a 31-Subagent org, refactored live down to 11
🎧 Episode 3 is live now.
https://open.spotify.com/episode/741mf2sJ0nFeM2JXAgWBP1?si=-lyvStsISbGfjzCswpePNw
https://youtu.be/LVvb2DE_zbw
@Salesforce Developer Group, Dubai, UAE @Agentblazer Community Group @* Salesforce Developers * @Architect Trailblazers @WFD Partner Cohort: TDX2630DaysChallenge
Good Morning Everyone,
I am a bit confused about the default Agentforce agent. I followed the steps in the Agentforce trailheads, but I am looking for more details on the types of agents.
- What is the core functionality behind the Agentforce (Default) Agent? I understand it should be enabled and activated, but what happens if I delete or deactivate it?
- I created a custom agent of type Service Agent from within Salesforce Sales/Service Cloud. However, it always opens the default Agentforce agent. Is there an option to mark the custom service agent as my default agent from inside Sales/Service Cloud? I deleted my default agent and now Agentforce is throwing an error.
- If we cant keep the custom agents as default agent, where and how the custom agent used ?
Any document or links appreciated.
Hi @Anoop Bhaskaran,
Currently I am also facing this issue, if you know the solution. Please let me know if it is really helpful for me.
Actually I want to connect my custom agents.
Thank you 🤝
📣 Closing the Agentforce Usage Visibility Gap
If leadership keeps asking "how much are we using Agentforce, and what's it costing?" — Salesforce just published a fix for that blind spot: the AI Agent Generative AI Usage Data Model.
It's a single DMO in Data 360 that refreshes every 5 minutes, capturing every metered and unmetered AI interaction with fields you can join to users, features, and CRM records.
🔗 Full breakdown: Close the Agentforce Usage Visibility Gap With the AI Usage Data Model
Trailhead Tuesday!
Discover how to blend digital accessibility and agentic AI to build truly inclusive experiences for everyone. 💡
From the badge, you'll learn:
- The essential intersection of digital accessibility and artificial intelligence
- How to address AI bias and navigate ethical challenges responsibly
- Strategies to design, test, and advocate for accessible AI products
Agentforce isn't magic. It's architecture and if you skip the architecture part, your agents will happily make very confident, very wrong decisions at 2 AM. No errors. No warnings. Just chaos with a smile.
In Episode 11, we open a brand-new season of The Franken Org focused entirely on Agentforce and agent architecture.
This episode covers:
- 🧠 What the Atlas Reasoning Engine actually does, step by step
- 🗂️ Subagents, Actions, and Instructions, the three levers you control
- 🛡️ Why grounding and the Trust Layer are non-negotiable
- 🔥 The (true-ish) story of the night FrankenForce's support agent approved 47 refunds before breakfast
🎧 Episode 11 is live now.
Youtube :
https://youtu.be/HTfTu1V45u8?si=p-o1XMTtXj8Vwn14
Spotify :
https://open.spotify.com/episode/5lu7894nDE2OmnFSav4Nu1
@Salesforce Developer Group, Dubai, UAE @Agentblazer Community Group @* Salesforce Developers * @WFD Partner Cohort: TDX2630DaysChallenge @Architect Trailblazers
Hi I created a prompt to have a summary of the activities of an account
I thought it would be easy with the instruction : Data to use :{!$RecordSnapshot:Account.Snapshot}
(I use the Field Generation template)
But in reality, it's not simple because I only got the future events and tasks.
I had to build a flow to get all activities.
Is it so, or did I miss something ?
Hi Hugues - you did not miss anything, that is expected. Activities are a special case: an Account surfaces them through TWO separate related lists - Open Activities (future events + open tasks) and Activity History (completed tasks + past events). The Account RecordSnapshot gives you the open/future set, so a summary grounded only on the snapshot naturally comes back with just upcoming events and open tasks, not the completed history you actually want to summarize.
So your Flow is the right call, not a hack. For a full activity summary you feed the data in yourself: a Flow (or Apex) that queries the Tasks and Events (or ActivityHistory + OpenActivity) filtered to the Account, then pass that collection into the prompt as a resource. That is the supported pattern - template-triggered prompt flows exist exactly for grounding on data the snapshot cannot give you.
Tip: cap and sort what the Flow returns (e.g. last N activities by date) so you stay within the prompt's context limit and the model summarizes the most relevant ones.
If this helps, please mark it as the Best Answer so it helps the next person - thanks :)