infiniflow / infiniflow/ragflow
[Feature Request]: /reset Command to Clear Conversation Context
- Dominant language
- Go
- Stars
- 91k
- Forks
- 10.8k
- Avg merge
- 1d 14h
- Merged PRs (30d)
- 705
Description
### Is there an existing issue for the same feature request?
- [x] I have checked the existing issues.
### Is your feature request related to a problem?
```Markdown
Problems to clear context in Third Apps via API
```
### Describe the feature you'd like
**Description:**
I would like to suggest adding a feature to clear the conversation context in a simple and effective way. The idea is to allow the user to send a command like `/reset` to delete the current conversation context, ensuring that the system starts with a fresh state, without relying on any previous information.
**Expected Behavior:**
- When the user sends the `/reset` command, the system should clear the conversation context and restart the interaction, not using any previously stored data.
**Example Usage:**
- User: `/reset`
- Expected response: "Conversation context reset. How can I assist you today?"
**Motivation:**
- This provides the user with more control over the interaction, especially when the model starts to hallucinate or generate nonsensical responses. Additionally, it is especially useful for applications that feed data through APIs (like Telegram and WhatsApp), where the system does not have direct access to the user's session, making it difficult to manage the conversation state.
### Describe implementation you've considered
_No response_
### Documentation, adoption, use case
```Markdown
```
### Additional information
_No response_
Contributor guide
No contributing guide indexed for this repository
Research direction
The issue specifies a /reset command and an expected fresh conversation response, but names no files, tests, or entry points. Trace how API messages are dispatched and where conversation context is stored, then define reset behavior and coverage for a subsequent interaction with no previous context.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- go
- Domain
- api, backend
- Issue type
- Feature
- Difficulty
- 5/5
- Estimated time
- Over a week
- Activity status
- Quiet
- Clarity
- Mostly clear
- Newbie friendliness
- 35/100