Error after updating Credentials for Confluence Loader

[wggrre]

Describe the bug

After changing the credentials used in a confluence document loader, the post/document-store/refresh/
{id} endpoint does not work anymore and result in a 403 (since the old invalid credentials are still beeing used).

The root cause is that the config object in the DB (table: document_store, column: loaders) is not updated correctly:

[
    {
        "id": "uuid....",
        "loaderId": "confluence",
        "loaderName": "SPACENAME",
        "loaderConfig": {
            "textSplitter": "",
            "baseUrl": "https://comapny.atlassian.net/wiki",
            "spaceKey": "SPACENAME",
            "limit": "",
            "metadata": "",
            "omitMetadataKeys": "",
            "FLOWISE_CREDENTIAL_ID": "new cred uuid",
            "credential": "new cred uuid"
        },
        "splitterId": "tokenTextSplitter",
        "splitterName": "Token Text Splitter",
        "splitterConfig": {
            "encodingName": "cl100k_base",
            "chunkSize": "1500",
            "chunkOverlap": "150"
        },
        "totalChunks": 296,
        "totalChars": 944389,
        "status": "SYNC",
        "credential": "old cred uuid",
        "source": "None"
    }
]

To Reproduce

1. Create a document store with an invalid credentials
2. Change to a new credentials -> the load is not working
3. Create a second loader use the new valid credentials -> loader ist working
4. compare the compare of the column loaders in table document_store

Expected behavior

When changing the credentials the config is still consistent and the endpoint still work
e.g.

[
    {
        "id": "uuid....",
        "loaderId": "confluence",
        "loaderName": "SPACENAME",
        "loaderConfig": {
            "textSplitter": "",
            "baseUrl": "https://comapny.atlassian.net/wiki",
            "spaceKey": "SPACENAME",
            "limit": "",
            "metadata": "",
            "omitMetadataKeys": "",
            "FLOWISE_CREDENTIAL_ID": "new cred uuid",
            "credential": "new cred uuid"
        },
        "splitterId": "tokenTextSplitter",
        "splitterName": "Token Text Splitter",
        "splitterConfig": {
            "encodingName": "cl100k_base",
            "chunkSize": "1500",
            "chunkOverlap": "150"
        },
        "totalChunks": 296,
        "totalChars": 944389,
        "status": "SYNC",
        "credential": "new cred uuid",
        "source": "None"
    }
]

Screenshots

No response

Flow

No response

Use Method

Docker

Flowise Version

3.1.2

Operating System

Linux

Browser

No response

Additional context

No response

[SyncWithRaj]

Hey! Thanks for the clear reproduction steps. I looked into this and found the root cause: the system was updating loaderConfig.credential correctly, but the top-level credential property was accidentally being overwritten with the stale DB value during the update process. This mismatch is what caused the downstream 403/404 errors during /refresh.

I've just opened a PR with a fix for this 🚀