Re: Flows or Apex I don't want to reinvent the wheel so was hoping there was somewhere that the existing QueryRecords invocable apex was accessible by us so I can then create my own starting with what they have. That action takes in a conversational prompt and translates that into a SOQL which is not something I wanted to reinvent if possible. I just need the LLM to do that I don't believe the records need to be fed to the LLM.
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.
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:
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:
🎧 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 :)
💙 Agents are making an impact at organizations!
Salesforce just published the second edition of the
Agentic Enterprise Index — and the numbers tell a pretty wild story about where the Agentic Enterprise is headed.
A few stats that jumped out:
The report tracks trends from February 2025 to April 2026 across real businesses actually using agents — not just talking about them.
🚀 5 Reasons to Move Your Agent to the New Agentforce Builder with Agent Script
Agent Script is here — and it's a big step up from the legacy Agentforce Builder. It combines deterministic logic with LLM reasoning (hybrid reasoning), giving you more predictable, reliable agent behavior — plus built-in observability, flexible build options (Canvas, Script, VS Code, CLI, or AI coding tools), and first access to all future Agentforce innovation like Voice and multi-agent orchestration.
✨ Good news — moving is straightforward, and both the new Agentforce Builder and Agent Script are included with your existing Agentforce purchase.
It’s alive... but is it scalable?
🧟♂️
Season 1 of The Franken Org Podcast: Technical Debt Autopsy is officially out in the wild!
If your org has a Slack thread from 2021 that just says "Ship it, we’ll fix it later"
(spoiler: nobody fixed it), 11 Apex triggers on a single object, and 418 unused fields... this season was recorded specifically for your therapeutic healing.
🎧 Listen/Watch now on your favorite platform:
💬 We need your feedback! Drop your thoughts, favorite anti-patterns, or worst org horror stories in the comments. Your feedback is critical for us as we prep Season 2
!
@Salesforce Developer Group, Dubai, UAE @Agentblazer Community Group @* Salesforce Developers * @WFD Partner Cohort: TDX2630DaysChallenge
Hi there,
i am looking for the insight on the following problem from people who successfully implemented and went live with the following scenario and may address the issues with verified in practice solutions.
we have implemented a service agent that must be exclusively grounded in the knowledge base with multiple subagents that share only one custom prompt action (now) that triggers the retriever (custom) with a related search index (vector search, with multilingual e5 large embedding model, on knowledge article object (data cloud). the agent is a service agent built in the new builder. prompt llm model is gemini 2.5 pro.
the knowledge base (standard sf knowledge solution) contains only 89 articles 9some very long, some shorter) but the information is law related and it is very complicated, it is all in English whereas the agent should support also the conversations in French, Spanish and Portuguese (as of now, we also need to roll out other languages after they stop being beta).
the subagents instructions and the prompt are also in english.
the problem we have (as of now there are 2 main levels):
1. agent wise: when the question (that was previously checked in english and returned a good answer) is asked in other languages the agent either does not understand the question at all or when it does understand it the answer comes back but is not satisfactory.
2. search index/retriever wise: when tested directly in the retriever we see that the scoring for the same chunks is lower significantly for other languages in comparison with the tested same question in english (open question if the scoring indeed passes together with chunks to LLM). to patch this we added keywords to knowledge articles in multiple languages, it helps but does not fix the problem.
the main proposed solutions (also by salesforce):
1. translate the knowledge base to all those languages we want the agent to speak (which still leaves open the fact that all the instructions and the prompt are in english).
2. normalize the query (user input) by translating it first to english and then query it while sending to prompt original language, original question and the translated english question. very simplistic description, understandable there is much more to it.
it feels that such behaviour must be handled totally out of the box and all proposed solutions are intuitive but feel an overkill for something that just should work.
is there anything i am missing? any insight will be very much appreciated.
Thank oyu
Yuliya
