[GH-ISSUE #24020] feat: Automations send to a single ongoing chat or per-automation ongoing chat instead of creating a new one every time #35685

New Issue

GiteaMirror · 2026-04-25T09:52:30-05:00

GiteaMirror commented

2026-04-25 09:52:30 -05:00

Originally created by @anathanscott on GitHub (Apr 22, 2026).
Original GitHub issue: https://github.com/open-webui/open-webui/issues/24020

Check Existing Issues

I have searched for all existing open AND closed issues and discussions for similar requests. I have found none that is comparable to my request.

Verify Feature Scope

I have read through and understood the scope definition for feature requests in the Issues section. I believe my feature request meets the definition and belongs in the Issues section instead of the Discussions.

Problem Description

Currently, Automations create a new chat every time they run. When managing a number of automations, this results in a massive glut of poorly-labeled chats that require manual cleanup, prevents models from having potentially relevant context to make decisions, and prevents features like TTS from firing on responses to automated prompts.

Desired Solution you'd like

We should have the ability to set all automations to fire into a single ongoing chat stream of our choice, or separate ongoing chats on a per-automation basis. These should be options alongside the current 'new chat every time' default. A simple setting like:

Create New Chat Every Run (current behavior, good for occasional generic automations that don't need context)
Single Chat for This Automation (creates a persistent chat for that specific automation, if user deletes, a fresh one is created on next run; good for frequent runs of an automation that need to stay organized in one place or need continual context)
Custom (select a specific pre-existing chat that all runs of the automation get passed to; ideal for assistant-style interactions where chat flow and broad context are critical)

Folder options would be great as well, for those who want new chats every run but don't want their interface getting cluttered.

Alternatives Considered

Currently, there are no alternatives. I have researched and played around extensively with ways to inject things into a specific ongoing chat, and it's quite complex to get working (multiple HTTP requests, merging messages with existing context, websocket to refresh chat state, etc). I've been unable to figure it out, and setting up for multiple tasks is a huge headache just to have your agent be prompted autonomously to issue a reminder or update in the context of an ongoing chat.

Additional Context

My usecase is a common one - a personal assistant agent that has tools to log and query health and fitness data, summarize emails, and manage a to-do list. Currently, I have a number of daily automations that would run covering similar things - check-ins to log health data, workout reminders, email summaries, etc. Currently, each new automation runs with zero context because it's in a new chat window, which can cause unpredictable agent behavior. It also prevents features like TTS from firing - I'm often away from my desk or have my hands full, and having my agent be able to autonomously summarize an email out loud or issue a verbal reminder would be a huge help. Finally, with my described usecase, I end every day with 10+ new chats with generic titles that need to be manually deleted one-by-one. For a user with more complex automations that run hourly or across multiple agents or users, the number of excess chats creates a poor and cluttered ux.

Originally created by @anathanscott on GitHub (Apr 22, 2026). Original GitHub issue: https://github.com/open-webui/open-webui/issues/24020 ### Check Existing Issues - [x] I have searched for all existing **open AND closed** issues and discussions for similar requests. I have found none that is comparable to my request. ### Verify Feature Scope - [x] I have read through and understood the scope definition for feature requests in the Issues section. I believe my feature request meets the definition and belongs in the Issues section instead of the Discussions. ### Problem Description Currently, Automations create a new chat every time they run. When managing a number of automations, this results in a massive glut of poorly-labeled chats that require manual cleanup, prevents models from having potentially relevant context to make decisions, and prevents features like TTS from firing on responses to automated prompts. ### Desired Solution you'd like We should have the ability to set all automations to fire into a single ongoing chat stream of our choice, or separate ongoing chats on a per-automation basis. These should be options alongside the current 'new chat every time' default. A simple setting like: **Create New Chat Every Run** (current behavior, good for occasional generic automations that don't need context) **Single Chat for This Automation** (creates a persistent chat for that specific automation, if user deletes, a fresh one is created on next run; good for frequent runs of an automation that need to stay organized in one place or need continual context) **Custom** (select a specific pre-existing chat that all runs of the automation get passed to; ideal for assistant-style interactions where chat flow and broad context are critical) Folder options would be great as well, for those who want new chats every run but don't want their interface getting cluttered. ### Alternatives Considered Currently, there are no alternatives. I have researched and played around extensively with ways to inject things into a specific ongoing chat, and it's quite complex to get working (multiple HTTP requests, merging messages with existing context, websocket to refresh chat state, etc). I've been unable to figure it out, and setting up for multiple tasks is a huge headache just to have your agent be prompted autonomously to issue a reminder or update in the context of an ongoing chat. ### Additional Context My usecase is a common one - a personal assistant agent that has tools to log and query health and fitness data, summarize emails, and manage a to-do list. Currently, I have a number of daily automations that would run covering similar things - check-ins to log health data, workout reminders, email summaries, etc. Currently, each new automation runs with zero context because it's in a new chat window, which can cause unpredictable agent behavior. It also prevents features like TTS from firing - I'm often away from my desk or have my hands full, and having my agent be able to autonomously summarize an email out loud or issue a verbal reminder would be a huge help. Finally, with my described usecase, I end every day with 10+ new chats with generic titles that need to be manually deleted one-by-one. For a user with more complex automations that run hourly or across multiple agents or users, the number of excess chats creates a poor and cluttered ux.

GiteaMirror closed this issue

2026-04-25 09:52:31 -05:00

GiteaMirror commented

2026-04-25 09:52:32 -05:00

@Lyhtande commented on GitHub (Apr 22, 2026):

Thanks for the request - I was just about to suggest that myself.
It would also be great if there were an option to send the automations to a specific channel.

@Lyhtande commented on GitHub (Apr 22, 2026): Thanks for the request - I was just about to suggest that myself. It would also be great if there were an option to send the automations to a specific channel.

GiteaMirror commented

2026-04-25 09:52:32 -05:00

@tjbck commented on GitHub (Apr 24, 2026):

#24074

@tjbck commented on GitHub (Apr 24, 2026): #24074

Sign in to join this conversation.

Branches Tags

1 Participants

Notifications

Due Date

No due date set.

Dependencies

No dependencies set.

Reference: github-starred/open-webui#35685