mirror of
https://github.com/open-webui/open-webui.git
synced 2026-05-08 21:09:41 -05:00
[GH-ISSUE #5962] enh: prompt management w/ version control #14186
No Assignees
Notifications
Due Date
No due date set.
Dependencies
No dependencies set.
Reference: github-starred/open-webui#14186
Reference in New Issue
Block a user
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?
Originally created by @flefevre on GitHub (Oct 7, 2024).
Original GitHub issue: https://github.com/open-webui/open-webui/issues/5962
Originally assigned to: @tjbck on GitHub.
Feature Request: Prompt Management System for Open WebUI
Is your feature request related to a problem? Please describe.
Managing prompts in Open WebUI can be cumbersome without a proper system. Hardcoded prompts lead to versioning challenges, complicate updates, and reduce flexibility for both technical and non-technical users. Currently, there is no efficient way to store, track, or manage prompts dynamically, limiting the ability to optimize or customize the system.
Describe the solution you'd like
I propose the implementation of a prompt management system within Open WebUI. This system would allow users to:
Additionally, this system should be capable of integrating with existing prompt management solutions to avoid reinventing the wheel. For example, by allowing connection to a third-party system, users can seamlessly leverage external prompt management tools without extensive modifications to Open WebUI.
Describe alternatives you've considered
Manually updating and managing prompts within the codebase is the current solution, which requires redeployments for changes and lacks efficient version control. This method is not scalable and adds unnecessary complexity.
Additional context
Integration with an external prompt management system (such as an API) would allow teams to use industry-standard tools for managing prompts, reducing development overhead while increasing flexibility and control.
@Capsar commented on GitHub (Dec 3, 2024):
+1 I am looking forward to this feature, it would be very helpful!
@superjamie commented on GitHub (Dec 26, 2024):
This would be a cool feature.
The Msty chat frontend has a similar functionality. It provides:
@flefevre commented on GitHub (Dec 26, 2024):
Ok, thanks for this link.
My initial idea was to be a 'prompt management system '
Key element were
Linking to langfuse could be also a good idea
@superjamie commented on GitHub (Dec 26, 2024):
If you wanted to pre-populate prompts, I found a set of CC0 prompts here:
https://github.com/f/awesome-chatgpt-prompts
This appears to be mostly what Msty uses as well.
@silentoplayz commented on GitHub (Aug 25, 2025):
Related - https://github.com/open-webui/open-webui/issues/3745
@tjbck commented on GitHub (Oct 2, 2025):
Soon.
@silentoplayz commented on GitHub (Oct 3, 2025):
Linking a closed PR of mine - https://github.com/open-webui/open-webui/pull/13991
@flefevre commented on GitHub (Oct 25, 2025):
I’d like to share some thoughts on the potential value of deeper Langfuse compatibility in Open WebUI — while fully acknowledging that this discussion also has a strategic dimension for the project’s roadmap.
I’ll align with whatever direction the Open WebUI team decides to take, but I think it’s worth considering the broader ecosystem benefits.
💡 Why Langfuse compatibility could be valuable
Prompt lifecycle and management
Langfuse offers mature tools for managing prompts — including versioning, rollback, history tracking, and branching.
These features help teams handle prompts as evolving assets rather than static strings, improving maintainability and experimentation.
Traceability and compliance
In some contexts (e.g. regulated industries), having full traceability of prompts — knowing which version was used, when and by whom — can be critical from a compliance or audit perspective.
Langfuse already provides this kind of prompt lineage functionality.
Ecosystem synergy
Strengthening the connection between Open WebUI and Langfuse wouldn’t just benefit one project — it could help reinforce the open-source ecosystem as a whole.
For example, Langfuse is already integrated with tools like LiteLLM for prompt management (docs.litellm.ai), showing that interoperability can unlock value across multiple components.
Strategic coexistence
I completely understand that Open WebUI needs to define its own identity and priorities.
My view is simply that having one strong open-source UI and a healthy, interconnected ecosystem around it can make the community as a whole more resilient and innovative.
🔧 Possible integration ideas (if aligned with the project’s vision)
In short: I see this not as a competing direction, but as a potentially complementary integration that could strengthen both Open WebUI and the broader open-source LLM tooling ecosystem.
Happy to align with the team’s strategy and contribute ideas if this direction becomes relevant.
@druellan commented on GitHub (Oct 25, 2025):
I'm right now learning prompting, and I'm already in the v5 of the small prompt I'm building for a project, and I'm feeling the need to proper versioning and testing. To have versioning in OWUI can be very helpful, specially for agents, where microadjusting a prompt is a thing, but I agree that proper integration with tools like Langfuse unlocks another level, with a proper lifecycle, testing and even orchestration.
OWUI also lacks a universal system prompt, so a single repository also helps to keep several models up-to-date.
@flefevre commented on GitHub (Nov 18, 2025):
I have continue to explore this key topic.
We need perhaps to look how Litellm is moving by integrating prompt in gitops.
They have created a binder to git repo with file extension .prompt
I think it will be key to have a standard between LiteLLM, langfuse and openwebui