open-webui/docs
1
0
Fork 0
mirror of https://github.com/open-webui/docs.git synced 2026-11-05 01:38:13 +00:00
Code Issues 17 Projects Packages Wiki Activity

[GH-ISSUE #1438] Backend-Controlled API Flow: user message disappears from UI after step 2 trigger assistant completion and refresh. #171

New issue
Closed
opened 2026-10-18 14:09:35 +00:00 by bsakaguchi · 2 comments
bsakaguchi commented 2026-10-18 14:09:35 +00:00
Owner
Copy link

Originally created by @vrzh on GitHub (Sep 23, 2026).
Original GitHub issue: https://github.com/open-webui/docs/issues/1438

Hi, thank you for the awesome project!

I am following the Backend-Controlled API Flow instructions. I hope this is the correct venue, there was a previous similar issue posted here https://github.com/open-webui/docs/issues/1175.

I understand that "Future versions may introduce changes in behavior or API structure." so I am hoping as part of the answer would be some info on how I can figure out what the correct calls are (if there are changes in the future).

When I complete step 2 (trigger assistant completion) and then refresh the web UI, the user message disappears from the UI. I am on the latest version 0.11.4.

UI view of chat after step 1 (as expected):
Image

UI view of chat after step 2 (as expected both user and assistant messages can be seen):

Image

UI view after refreshing the UI in the browser (problem - only the assistant message can be seen, user message is missing):
Image

For ease of testing I combined step 1 and step 2 into a script:

API_KEY="..."
HOST="..."

# Step 1: Create Chat with User and Assistant Messages
# https://docs.openwebui.com/reference/api-flow/#step-1-create-chat-with-user-and-assistant-messages

USER_MSG_ID=$(uuidgen || python3 -c "import uuid; print(uuid.uuid4())")
ASSISTANT_MSG_ID=$(uuidgen || python3 -c "import uuid; print(uuid.uuid4())")
TIMESTAMP=$(date +%s)

CHAT_ID=$(curl -X POST $HOST/api/v1/chats/new \
  -H "Authorization: Bearer $API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "chat": {
      "title": "New Chat",
      "models": ["gpt-5.6-sol"],
      "messages": [
        {
          "id": "'"$USER_MSG_ID"'",
          "role": "user",
          "content": "Hi, what is the capital of France?",
          "timestamp": '"$TIMESTAMP"',
          "models": ["gpt-5.6-sol"],
          "childrenIds": ["'"$ASSISTANT_MSG_ID"'"]
        },
        {
          "id": "'"$ASSISTANT_MSG_ID"'",
          "role": "assistant",
          "content": "",
          "parentId": "'"$USER_MSG_ID"'",
          "childrenIds": [],
          "model": "gpt-5.6-sol",
          "modelName": "gpt-5.6-sol",
          "modelIdx": 0,
          "done": false,
          "timestamp": '"$((TIMESTAMP + 1))"'
        }
      ],
      "history": {
        "currentId": "'"$ASSISTANT_MSG_ID"'",
        "messages": {
          "'"$USER_MSG_ID"'": {
            "id": "'"$USER_MSG_ID"'",
            "role": "user",
            "content": "Hi, what is the capital of France?",
            "timestamp": '"$TIMESTAMP"',
            "models": ["gpt-5.6-sol"],
            "childrenIds": ["'"$ASSISTANT_MSG_ID"'"]
          },
          "'"$ASSISTANT_MSG_ID"'": {
            "id": "'"$ASSISTANT_MSG_ID"'",
            "role": "assistant",
            "content": "",
            "parentId": "'"$USER_MSG_ID"'",
            "childrenIds": [],
            "model": "gpt-5.6-sol",
            "modelName": "gpt-5.6-sol",
            "modelIdx": 0,
            "done": false,
            "timestamp": '"$((TIMESTAMP + 1))"'
          }
        }
      }
    }
  }' | jq -r '.id')

echo "$CHAT_ID"


sleep 30

# Step 2: Trigger Assistant Completion
# https://docs.openwebui.com/reference/api-flow/#step-2-trigger-assistant-completion


SESSION_ID=$(uuidgen || python3 -c "import uuid; print(uuid.uuid4())")

curl -X POST $HOST/api/chat/completions \
  -H "Authorization: Bearer $API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "chat_id": "'"$CHAT_ID"'",
    "id": "'"$ASSISTANT_MSG_ID"'",
    "messages": [
      {
        "role": "user",
        "content": "Hi, what is the capital of France?"
      }
    ],
    "model": "gpt-5.6-sol",
    "stream": true,
    "background_tasks": {
      "title_generation": true,
      "tags_generation": false,
      "follow_up_generation": false
    },
    "features": {
      "code_interpreter": false,
      "web_search": false,
      "image_generation": false,
      "memory": false
    },
    "variables": {
      "{{USER_NAME}}": "",
      "{{USER_LANGUAGE}}": "en-US",
      "{{CURRENT_DATETIME}}": "2025-07-14T12:00:00Z",
      "{{CURRENT_TIMEZONE}}": "Europe"
    },
    "session_id": "'"$SESSION_ID"'"
  }'
Originally created by @vrzh on GitHub (Sep 23, 2026). Original GitHub issue: https://github.com/open-webui/docs/issues/1438 Hi, thank you for the awesome project! I am following the [Backend-Controlled API Flow](https://docs.openwebui.com/reference/api-flow/#step-2-trigger-assistant-completion) instructions. I hope this is the correct venue, there was a previous similar issue posted here https://github.com/open-webui/docs/issues/1175. I understand that "Future versions may introduce changes in behavior or API structure." so I am hoping as part of the answer would be some info on how I can figure out what the correct calls are (if there are changes in the future). When I complete step 2 (trigger assistant completion) and then refresh the web UI, the user message disappears from the UI. I am on the latest version `0.11.4`. UI view of chat after step 1 (as expected): <img width="806" height="240" alt="Image" src="https://github.com/user-attachments/assets/a2af892f-c2ad-493f-8317-a7f4226eb5f7" /> UI view of chat after step 2 (as expected both user and assistant messages can be seen): <img width="937" height="291" alt="Image" src="https://github.com/user-attachments/assets/ad85e4ad-063f-4d54-b3ef-4eb6e0331472" /> UI view after refreshing the UI in the browser (problem - only the assistant message can be seen, user message is missing): <img width="943" height="225" alt="Image" src="https://github.com/user-attachments/assets/c25c1af2-cc87-4f02-a353-fe005059cb57" /> For ease of testing I combined step 1 and step 2 into a script: ```sh API_KEY="..." HOST="..." # Step 1: Create Chat with User and Assistant Messages # https://docs.openwebui.com/reference/api-flow/#step-1-create-chat-with-user-and-assistant-messages USER_MSG_ID=$(uuidgen || python3 -c "import uuid; print(uuid.uuid4())") ASSISTANT_MSG_ID=$(uuidgen || python3 -c "import uuid; print(uuid.uuid4())") TIMESTAMP=$(date +%s) CHAT_ID=$(curl -X POST $HOST/api/v1/chats/new \ -H "Authorization: Bearer $API_KEY" \ -H "Content-Type: application/json" \ -d '{ "chat": { "title": "New Chat", "models": ["gpt-5.6-sol"], "messages": [ { "id": "'"$USER_MSG_ID"'", "role": "user", "content": "Hi, what is the capital of France?", "timestamp": '"$TIMESTAMP"', "models": ["gpt-5.6-sol"], "childrenIds": ["'"$ASSISTANT_MSG_ID"'"] }, { "id": "'"$ASSISTANT_MSG_ID"'", "role": "assistant", "content": "", "parentId": "'"$USER_MSG_ID"'", "childrenIds": [], "model": "gpt-5.6-sol", "modelName": "gpt-5.6-sol", "modelIdx": 0, "done": false, "timestamp": '"$((TIMESTAMP + 1))"' } ], "history": { "currentId": "'"$ASSISTANT_MSG_ID"'", "messages": { "'"$USER_MSG_ID"'": { "id": "'"$USER_MSG_ID"'", "role": "user", "content": "Hi, what is the capital of France?", "timestamp": '"$TIMESTAMP"', "models": ["gpt-5.6-sol"], "childrenIds": ["'"$ASSISTANT_MSG_ID"'"] }, "'"$ASSISTANT_MSG_ID"'": { "id": "'"$ASSISTANT_MSG_ID"'", "role": "assistant", "content": "", "parentId": "'"$USER_MSG_ID"'", "childrenIds": [], "model": "gpt-5.6-sol", "modelName": "gpt-5.6-sol", "modelIdx": 0, "done": false, "timestamp": '"$((TIMESTAMP + 1))"' } } } } }' | jq -r '.id') echo "$CHAT_ID" sleep 30 # Step 2: Trigger Assistant Completion # https://docs.openwebui.com/reference/api-flow/#step-2-trigger-assistant-completion SESSION_ID=$(uuidgen || python3 -c "import uuid; print(uuid.uuid4())") curl -X POST $HOST/api/chat/completions \ -H "Authorization: Bearer $API_KEY" \ -H "Content-Type: application/json" \ -d '{ "chat_id": "'"$CHAT_ID"'", "id": "'"$ASSISTANT_MSG_ID"'", "messages": [ { "role": "user", "content": "Hi, what is the capital of France?" } ], "model": "gpt-5.6-sol", "stream": true, "background_tasks": { "title_generation": true, "tags_generation": false, "follow_up_generation": false }, "features": { "code_interpreter": false, "web_search": false, "image_generation": false, "memory": false }, "variables": { "{{USER_NAME}}": "", "{{USER_LANGUAGE}}": "en-US", "{{CURRENT_DATETIME}}": "2025-07-14T12:00:00Z", "{{CURRENT_TIMEZONE}}": "Europe" }, "session_id": "'"$SESSION_ID"'" }' ```
bsakaguchi commented 2026-10-18 14:09:36 +00:00
Author
Owner
Copy link

@Classic298 commented on GitHub (Sep 23, 2026):

Thanks for the clean repro. The docs example was missing a field: the completion request needs user_message, the user message object you created in step 1. Without it the backend saves the assistant reply with no parent, so after a reload the UI can't find your message anymore.

Add this to the step 2 request and it works:

"user_message": {
  "id": "<USER_MSG_ID>",
  "role": "user",
  "content": "Hi, what is the capital of France?",
  "parentId": null,
  "childrenIds": ["<ASSISTANT_MSG_ID>"],
  "models": ["gpt-5.6-sol"]
}

Docs are updated. For future changes: send a message in the web UI with the browser dev tools open and look at the /api/chat/completions request in the Network tab. That body is exactly what the UI sends, so it's always the current reference.

<!-- gh-comment-id:5795574432 --> @Classic298 commented on GitHub (Sep 23, 2026): Thanks for the clean repro. The docs example was missing a field: the completion request needs `user_message`, the user message object you created in step 1. Without it the backend saves the assistant reply with no parent, so after a reload the UI can't find your message anymore. Add this to the step 2 request and it works: ```json "user_message": { "id": "<USER_MSG_ID>", "role": "user", "content": "Hi, what is the capital of France?", "parentId": null, "childrenIds": ["<ASSISTANT_MSG_ID>"], "models": ["gpt-5.6-sol"] } ``` Docs are updated. For future changes: send a message in the web UI with the browser dev tools open and look at the `/api/chat/completions` request in the Network tab. That body is exactly what the UI sends, so it's always the current reference.
bsakaguchi commented 2026-10-18 14:09:36 +00:00
Author
Owner
Copy link

@vrzh commented on GitHub (Sep 23, 2026):

Thank you, works like a charm!

I initially did look at the dev tools, but saw only a request to completions (i.e. no step 1 request):

Image

This got me confused, I thought it did not apply. But now I know!

<!-- gh-comment-id:5795804173 --> @vrzh commented on GitHub (Sep 23, 2026): Thank you, works like a charm! I initially did look at the dev tools, but saw only a request to `completions` (i.e. no step 1 request): <img width="257" height="170" alt="Image" src="https://github.com/user-attachments/assets/cff3764d-a699-43ff-a577-f6de56fd2045" /> This got me confused, I thought it did not apply. But now I know!
Sign in to join this conversation.
No Branch/Tag specified
Branches Tags
main
dev
No results found.
Labels
Clear labels   
bug

   
documentation

   
documentation

   
duplicate

   
enhancement

   
enhancement

   
good first issue

   
good first issue

   
good first issue

   
help wanted

   
help wanted

   
pull-request
Mirrored from GitHub Pull Request

No labels
bug
documentation
documentation
duplicate
enhancement
enhancement
good first issue
good first issue
good first issue
help wanted
help wanted
pull-request
Milestone
Clear milestone
No items
No milestone
Projects
Clear projects
No items
No project
Assignees
Clear assignees
No assignees
1 participant
Notifications
Due date
The due date is invalid or out of range. Please use the format "yyyy-mm-dd".

No due date set.

Dependencies

No dependencies set.

Reference
open-webui/docs#171
No description provided.