Mintplex-Labs/anything-llm
1
0
Fork 0
mirror of https://github.com/Mintplex-Labs/anything-llm.git synced 2026-07-21 01:25:28 -04:00
Code Issues 328 Packages Projects Releases 10 Wiki Activity

[GH-ISSUE #4229] [FEAT]: Follow-up button, which allowed to quickly ask questions about a specific section #2689

New Issue
Open
opened 2026-02-22 18:30:48 -05:00 by yindo · 5 comments
yindo commented 2026-02-22 18:30:48 -05:00
Owner
Copy Link
Copy Source

Originally created by @roman-y-wu on GitHub (Jul 31, 2025).
Original GitHub issue: https://github.com/Mintplex-Labs/anything-llm/issues/4229

What would you like to see?

It handles conversational continuity—in other words, its ability to:

  • Understand prior context
  • Respond intelligently to clarifying or extension questions
  • Chain reasoning or exploration across multiple turns
Originally created by @roman-y-wu on GitHub (Jul 31, 2025). Original GitHub issue: https://github.com/Mintplex-Labs/anything-llm/issues/4229 ### What would you like to see? It handles conversational continuity—in other words, its ability to: - Understand prior context - Respond intelligently to clarifying or extension questions - Chain reasoning or exploration across multiple turns
yindo added the enhancementfeature request labels 2026-02-22 18:30:48 -05:00
yindo commented 2026-02-22 18:30:49 -05:00
Author
Owner
Copy Link
Copy Source

@therealtimex commented on GitHub (Jul 31, 2025):

What the UI/UX looks like?

@therealtimex commented on GitHub (Jul 31, 2025): What the UI/UX looks like?
yindo commented 2026-02-22 18:30:49 -05:00
Author
Owner
Copy Link
Copy Source

@timothycarambat commented on GitHub (Jul 31, 2025):

The core issue with features like this, while they do improve the UI/UX is the wasted cycles to make the content. Some people really struggle to run even a 3B on device. Tasking the LLM to, for example, go back over the conversation and make 3 follows up or make a clever thread title, is taking up memory on the client device without them explictly asking for it.

For local devices where hardware is limited, this results in laggy performance since the model is doing something else in the background. If connected to cloud, this results in additional spend.

Not to say this is not a good or bad idea, but we always need to keep in mind the person on an old Lenovo laptop running a 3B model on CPU. We do have to draw a line somewhere though! Its a balancing act.

Personally, i like the idea, maybe its opt-in or something.

@timothycarambat commented on GitHub (Jul 31, 2025): The core issue with features like this, while they do improve the UI/UX is the wasted cycles to make the content. Some people _really_ struggle to run even a 3B on device. Tasking the LLM to, for example, go back over the conversation and make 3 follows up or make a clever thread title, is taking up memory on the client device without them explictly asking for it. For local devices where hardware is limited, this results in laggy performance since the model is doing something else in the background. If connected to cloud, this results in additional spend. Not to say this is not a good or bad idea, but we always need to keep in mind the person on an old Lenovo laptop running a 3B model on CPU. We do have to draw a line somewhere though! Its a balancing act. Personally, i like the idea, maybe its opt-in or something.
yindo commented 2026-02-22 18:30:49 -05:00
Author
Owner
Copy Link
Copy Source

@DeightonLLM commented on GitHub (Aug 2, 2025):

Maybe a middle ground like a toggle switch. We miss the ability to use that feature when drilling down into a topic. We do use do mainly use API calls though.

L

@DeightonLLM commented on GitHub (Aug 2, 2025): Maybe a middle ground like a toggle switch. We miss the ability to use that feature when drilling down into a topic. We do use do mainly use API calls though. L
yindo commented 2026-02-22 18:30:49 -05:00
Author
Owner
Copy Link
Copy Source

@Alminc91 commented on GitHub (Oct 2, 2025):

@roman-y-wu @timothycarambat
What I do is I have specified in my system prompt for the workspace that the LLM should output a topic-specific follow-up question at the end.

This is really nice and makes the bot more conversational (although without conversational RAG this feature can lead to wrong outputs). Making it clickable would be especially useful for the embed chatbot. The Botsonic chatbot has this really nicely integrated.

If the follow-up would be clickable, that is quite user-friendly. Maybe a possibility is to use one LLM cycle, if possible. Like with my system prompt scenario, where it creates one at the end on-the-go. I wonder how Botsonic does it.

@Alminc91 commented on GitHub (Oct 2, 2025): @roman-y-wu @timothycarambat What I do is I have specified in my system prompt for the workspace that the LLM should output a topic-specific follow-up question at the end. This is really nice and makes the bot more conversational (although without conversational RAG this feature can lead to wrong outputs). Making it clickable would be especially useful for the embed chatbot. The Botsonic chatbot has this really nicely integrated. If the follow-up would be clickable, that is quite user-friendly. Maybe a possibility is to use one LLM cycle, if possible. Like with my system prompt scenario, where it creates one at the end on-the-go. I wonder how Botsonic does it.
yindo commented 2026-02-22 18:30:49 -05:00
Author
Owner
Copy Link
Copy Source

@timothycarambat commented on GitHub (Oct 2, 2025):

We could make this an optional default skill so that it only wastes compute cycles if you want it to. Something along those lines is probably a good fit because then we can specifically highlight them as buttons and not just text.

Botsonic does this because they only use fast cloud-based LLMs and simply just eat the cost for the UI/UX of this, either doing it per prompt or on document/vector change.

@timothycarambat commented on GitHub (Oct 2, 2025): We could make this an optional default skill so that it only wastes compute cycles if you want it to. Something along those lines is probably a good fit because then we can specifically highlight them as buttons and not just text. Botsonic does this because they only use fast cloud-based LLMs and simply just eat the cost for the UI/UX of this, either doing it per prompt or on document/vector change.
yindo changed title from [FEAT]: Follow-up button, which allowed to quickly ask questions about a specific section to [GH-ISSUE #4229] [FEAT]: Follow-up button, which allowed to quickly ask questions about a specific section 2026-06-05 14:47:59 -04:00
Sign in to join this conversation.
No Branch/Tag Specified
Branches Tags
Labels
Clear labels
Desktop
Docker
Integration Request
Integration Request
OS: Linux
OS: Mobile
OS: Windows
UI/UX
blocked
bug
bug
core-team-only
documentation
duplicate
embed-widget
enhancement
feature request
github_actions
good first issue
investigating
needs info / can't replicate
possible bug
pull-request
Mirrored from GitHub Pull Request
 question
stage: specifications
wontfix
No Label enhancement feature request
Milestone
No items
No Milestone
Projects
Clear projects
No project
Assignees
Clear assignees
yindo
No Assignees
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: Mintplex-Labs/anything-llm#2689
Delete Branch "%!s()"

Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?