### What problem does this PR solve?

Hello, our use case requires LLM agent to invoke some tools, so I made a
simple implementation here.

This PR does two things:

1. A simple plugin mechanism based on `pluginlib`:

This mechanism lives in the `plugin` directory. It will only load
plugins from `plugin/embedded_plugins` for now.

A sample plugin `bad_calculator.py` is placed in
`plugin/embedded_plugins/llm_tools`, it accepts two numbers `a` and `b`,
then give a wrong result `a + b + 100`.

In the future, it can load plugins from external location with little
code change.

Plugins are divided into different types. The only plugin type supported
in this PR is `llm_tools`, which must implement the `LLMToolPlugin`
class in the `plugin/llm_tool_plugin.py`.
More plugin types can be added in the future.

2. A tool selector in the `Generate` component:

Added a tool selector to select one or more tools for LLM:


![image](https://github.com/user-attachments/assets/74a21fdf-9333-4175-991b-43df6524c5dc)

And with the `bad_calculator` tool, it results this with the `qwen-max`
model:


![image](https://github.com/user-attachments/assets/93aff9c4-8550-414a-90a2-1a15a5249d94)


### Type of change

- [ ] Bug Fix (non-breaking change which fixes an issue)
- [x] New Feature (non-breaking change which adds functionality)
- [ ] Documentation Update
- [ ] Refactoring
- [ ] Performance Improvement
- [ ] Other (please describe):

Co-authored-by: Yingfeng <yingfeng.zhang@gmail.com>

### What problem does this PR solve?

Improve oauth configuration documentation and examples.

- Related pull requests: 
  - #7379
  - #7553
  - #7587
- Related issues:
  -  #3495
### Type of change

- [x] Documentation Update

### What problem does this PR solve?

Launch sandbox from docker-compose.
#4977
### Type of change

- [x] New Feature (non-breaking change which adds functionality)
- [x] Documentation Update

---------

Co-authored-by: writinwaters <93570324+writinwaters@users.noreply.github.com>

### What problem does this PR solve?

Close #7655

Based on the codes atthe api_app, I think the reference is one-to-one
with the message
`
    def fillin_conv(ans):
        nonlocal conv, message_id
        if not conv.reference:
            conv.reference.append(ans["reference"])
        else:
            conv.reference[-1] = ans["reference"]
conv.message[-1] = {"role": "assistant", "content": ans["answer"], "id":
message_id}
        ans["id"] = message_id
`



### Type of change
- [x] Bug Fix (non-breaking change which fixes an issue)

### What problem does this PR solve?

This PR introduces Pydantic-based validation for the delete dataset HTTP
API, improving code clarity and robustness. Key changes include:

1. Pydantic Validation
2. Error Handling
3. Test Updates
4. Documentation Updates

### Type of change

- [x] Documentation Update
- [x] Refactoring

### What problem does this PR solve?

Deprecate `/github_callback` route in favor of
`/oauth/callback/<channel>` for GitHub OAuth integration:

- Added GitHub OAuth support in the authentication module
- Introduced `GithubOAuthClient` with methods to fetch and normalize
user info
  - Updated `CLIENT_TYPES` to include GitHub OAuth client
- Deprecated `/github_callback` route and suggested using the generic
`/oauth/callback/<channel>` route

---
- Related pull requests: 
  - #7379
  - #7553 

### Usage

- [Create a GitHub OAuth
App](https://github.com/settings/applications/new) to obtain the
`client_id` and `client_secret`, configure the authorization callback
url: `https://your-app.com/v1/user/oauth/callback/github`
- Edit `service_conf.yaml.template`:
  ```yaml
  # ...
  oauth:
    github:
      type: "github"
      icon: "github"
      display_name: "Github"
      client_id: "your_client_id"
      client_secret: "your_client_secret"
      redirect_uri: "https://your-app.com/v1/user/oauth/callback/github"
  # ...
  ```

### Type of change

- [x] Documentation Update
- [x] Refactoring (non-breaking change)

### What problem does this PR solve?

When Delete Chunk Will Also Delete Chunk Related Image

### Type of change
- [x] New Feature (non-breaking change which adds functionality)

### What problem does this PR solve?

#7623

### Type of change

- [x] Bug Fix (non-breaking change which fixes an issue)

### What problem does this PR solve?

Close #7612

### Type of change

- [x] Bug Fix (non-breaking change which fixes an issue)

### What problem does this PR solve?

Close #7597

### Type of change

- [x] Bug Fix (non-breaking change which fixes an issue)

### What problem does this PR solve?

#7623

### Type of change

- [x] New Feature (non-breaking change which adds functionality)

### What problem does this PR solve?

Close #7592

### Type of change

- [x] Bug Fix (non-breaking change which fixes an issue)

### What problem does this PR solve?

Fix  HTTP API Create/Update dataset parser config default value error

### Type of change

- [x] Bug Fix (non-breaking change which fixes an issue)

### What problem does this PR solve?

Hello, we are using ragflow as a backend service, so we need to manage
agents from our own frontend. So adding these http APIs to manage
agents.

The code logic is copied and modified from the `rm` and `save` methods
in `api/apps/canvas_app.py`.

btw, I found that the `save` method in `canvas_app.py` actually allows
to modify an agent to an existing title, so I kept the behavior in the
http api. I'm not sure if this is intentional.

### Type of change

- [ ] Bug Fix (non-breaking change which fixes an issue)
- [x] New Feature (non-breaking change which adds functionality)
- [x] Documentation Update
- [ ] Refactoring
- [ ] Performance Improvement
- [ ] Other (please describe):

### What problem does this PR solve?

add show debug
![Recording2025-04-28142829-ezgif
com-video-to-gif-converter](https://github.com/user-attachments/assets/0c67da34-c2b6-428f-ae9b-b5b21464885c)

### Type of change

- [x] New Feature (non-breaking change which adds functionality)

### What problem does this PR solve?


Close #7567

### Type of change

- [x] Bug Fix (non-breaking change which fixes an issue)

### What problem does this PR solve?

This PR introduces Pydantic-based validation for the update dataset HTTP
API, improving code clarity and robustness. Key changes include:
1. Pydantic Validation
2. ​​Error Handling
3. Test Updates
4. Documentation Updates
5. fix bug: #5915

### Type of change

- [x] Bug Fix (non-breaking change which fixes an issue)
- [x] Documentation Update
- [x] Refactoring

### What problem does this PR solve?

1. The MySQL instance is configured with max_connections=1000,
but our connection pool was limited to max_connections: 100.
This mismatch caused connection pool exhaustion during performance
testing.

2.  Increase stale_timeout to resolve #6548

### Type of change

- [x] Performance Improvement

### What problem does this PR solve?

#7376
#4503
#5710 
#7470

### Type of change

- [x] New Feature (non-breaking change which adds functionality)

### What problem does this PR solve?

_Briefly describe what this PR aims to solve. Include background context
that will help reviewers understand the purpose of the PR._

### Type of change

- [ ] Bug Fix (non-breaking change which fixes an issue)
- [ ] New Feature (non-breaking change which adds functionality)
- [ ] Documentation Update
- [ ] Refactoring
- [ ] Performance Improvement
- [ ] Other (please describe):

### What problem does this PR solve?

Kb detail supports return document total size now.

### Type of change

- [x] New Feature (non-breaking change which adds functionality)

### What problem does this PR solve?
https://github.com/infiniflow/ragflow/issues/7519
### Type of change
- [x] New Feature (non-breaking change which adds functionality)

---------

Co-authored-by: Kevin Hu <kevinhu.sh@gmail.com>

### What problem does this PR solve?

Add `/login/channels` route and improve auth logic to support frontend
integration with third-party login providers:

- Add `/login/channels` route to provide authentication channel list
with `display_name` and `icon`
- Optimize user info parsing logic by prioritizing `avatar_url` and
falling back to `picture`
- Simplify OIDC token validation by removing unnecessary `kid` checks
- Ensure `client_id` is safely cast to string during `audience`
validation
- Fix typo

---
- Related pull request: #7379 

### Type of change

- [x] New Feature (non-breaking change which adds functionality)
- [x] Documentation Update

### What problem does this PR solve?

#7484

### Type of change

- [x] Bug Fix (non-breaking change which fixes an issue)

Fix:When sharing the knowledge base of multiple tenants with one person,
when this person queries the knowledge base of both tenants, they will
only query the question of the first person's knowledge base

Co-authored-by: 杜有强 <duyq@internal.ths.com.cn>

### What problem does this PR solve?

1.  Add delete_by_ids method
2. Add get_doc_ids_by_doc_names
3. Improve user_canvan_version's logic (avoid O(n) db IO)
4. Improve document delete logic (avoid O(n) db IO)

### Type of change

- [x] Performance Improvement

### What problem does this PR solve?

Fix: After deleting the file from the file management menu, it was not
removed from the MinIO bucket.

### Type of change

- [x] Bug Fix (non-breaking change which fixes an issue)
- [ ] New Feature (non-breaking change which adds functionality)
- [ ] Documentation Update
- [ ] Refactoring
- [ ] Performance Improvement
- [ ] Other (please describe):

Co-authored-by: liuzhenghua-jk <liuzhenghua-jk@360shuke.com>

### What problem does this PR solve?



### Type of change

- [x] Bug Fix (non-breaking change which fixes an issue)

### What problem does this PR solve?

Optimize dataset validation and add function docs

### Type of change

- [x] Refactoring

### What problem does this PR solve?

Fix `filed_map` was incorrectly persisted. #7412 

### Type of change

- [x] Bug Fix (non-breaking change which fixes an issue)

### What problem does this PR solve?

change create dataset delimiter default value to r'\n'

### Type of change

- [x] New Feature (non-breaking change which adds functionality)

### What problem does this PR solve?

https://github.com/infiniflow/ragflow/issues/7407

Based on this context, I think there should be some reasons that let
some LLMs have a mismatch (add the wrong "@xxx"),
So I think when use fid can not fetch llm then tried to just use name
should can fetch it.

### Type of change

- [x] Bug Fix (non-breaking change which fixes an issue)

### What problem does this PR solve?

Remove unnecessary parameter restrictions in dataset creation API

### Type of change

- [x] Bug Fix (non-breaking change which fixes an issue)

### What problem does this PR solve?

When you removed any document in a knowledge base using knowledge graph,
the graph's `removed_kwd` is set to "Y".
However, in the function `graphrag.utils.get_gaph`, `rebuild_graph`
method is passed and directly return `None` while `removed_kwd=Y`,
making residual part of the graph abandoned (but old entity data still
exist in db).

Besides, infinity instance actually pass deleting graph components'
`source_id` when removing document. It may cause wrong graph after
rebuild.

### Type of change

- [x] Bug Fix (non-breaking change which fixes an issue)

### What problem does this PR solve?

This PR introduces Pydantic-based validation for the create dataset HTTP
API, improving code clarity and robustness. Key changes include:
1. Pydantic Validation
2. ​​Error Handling
3. Test Updates
4. Documentation

### Type of change

- [x] Bug Fix (non-breaking change which fixes an issue)
- [x] Documentation Update
- [x] Refactoring

### What problem does this PR solve?

Add support for OAuth2 and OpenID Connect (OIDC) authentication,
allowing OAuth/OIDC authentication using the specified routes:
- `/login/<channel>`: Initiates the OAuth flow for the specified channel
- `/oauth/callback/<channel>`: Handles the OAuth callback after
successful authentication

The callback URL should be configured in your OAuth provider as:
```
https://your-app.com/oauth/callback/<channel>
```

For detailed instructions on configuring **service_conf.yaml.template**,
see: `./api/apps/auth/README.md#usage`.

- Related issues
#3495  

### Type of change

- [x] New Feature (non-breaking change which adds functionality)
- [x] Documentation Update

### What problem does this PR solve?

When updating a chat assistant using API,if the dataset attached by the
current chat assistant is not empty，setting dataset to
null("dataset_ids":[]) will cause update failure：'dataset_ids' can't be
empty

### Type of change

- [x] Bug Fix (non-breaking change which fixes an issue)

### What problem does this PR solve?

In the generate_confirmation_token method, a spelling error was found
with 'tenent_id'. The correct spelling should be 'tenant_id'.

### Type of change

- [x] Bug Fix (non-breaking change which fixes an issue)
- [ ] New Feature (non-breaking change which adds functionality)
- [ ] Documentation Update
- [ ] Refactoring
- [ ] Performance Improvement
- [ ] Other (please describe):

Co-authored-by: shengliang xiao <shengliangxiao2024@gmail.com>

### What problem does this PR solve?

Enhance capability of `list_docs`.

Breaking change: change method from `GET` to `POST`.

### Type of change

- [x] Refactoring
- [x] Enhancement with breaking change

### What problem does this PR solve?



### Type of change


- [x] Documentation Update

### What problem does this PR solve?

According to the
[[Rucongzhang](https://github.com/Rucongzhang)](https://github.com/infiniflow/ragflow/pull/7057#issuecomment-2827410047)
I added DB reconnection strategy in function `update_by_id`

### What problem does this PR solve?

This PR adds the support for latest OpenSearch2.19.1 as the store engine
& search engine option for RAGFlow.

### Main Benefit

1. OpenSearch2.19.1 is licensed under the [Apache v2.0 License] which is
much better than Elasticsearch
2. For search, OpenSearch2.19.1 supports full-text
search、vector_search、hybrid_search those are similar with Elasticsearch
on schema
3. For store, OpenSearch2.19.1 stores text、vector those are quite
simliar with Elasticsearch on schema

### Changes

- Support opensearch_python_connetor. I make a lot of adaptions since
the schema and api/method between ES and Opensearch differs in many
ways(especially the knn_search has a significant gap) :
rag/utils/opensearch_coon.py
- Support static config adaptions by changing:
conf/service_conf.yaml、api/settings.py、rag/settings.py
- Supprt some store&search schema changes between OpenSearch and ES:
conf/os_mapping.json
- Support OpenSearch python sdk : pyproject.toml
- Support docker config for OpenSearch2.19.1 :
docker/.env、docker/docker-compose-base.yml、docker/service_conf.yaml.template

### How to use
- I didn't change the priority that ES as the default doc/search engine.
Only if in docker/.env , we set DOC_ENGINE=${DOC_ENGINE:-opensearch}, it
will work.


### Others
Our team tested a lot of docs in our environment by using OpenSearch as
the vector database ,it works very well.
All the conifg for OpenSearch is necessary.

### Type of change

- [x] New Feature (non-breaking change which adds functionality)

---------

Co-authored-by: Yongteng Lei <yongtengrey@outlook.com>
Co-authored-by: writinwaters <93570324+writinwaters@users.noreply.github.com>
Co-authored-by: Yingfeng <yingfeng.zhang@gmail.com>

### What problem does this PR solve?

Some models force thinking, resulting in the absence of the think tag in
the returned content

### Type of change

- [x] Bug Fix (non-breaking change which fixes an issue)

### What problem does this PR solve?

Get the highlight parameter from the request to keep consistency with
the document

> 
- Method: POST
- URL: `/api/v1/retrieval`
- Headers:
  - `'content-Type: application/json'`
  - `'Authorization: Bearer <YOUR_API_KEY>'`
- Body:
  - `"question"`: `string`  
  - `"dataset_ids"`: `list[string]`  
  - `"document_ids"`: `list[string]`
  - `"page"`: `integer`  
  - `"page_size"`: `integer`  
  - `"similarity_threshold"`: `float`  
  - `"vector_similarity_weight"`: `float`  
  - `"top_k"`: `integer`  
  - `"rerank_id"`: `string`  
  - `"keyword"`: `boolean`  
  - `"highlight"`: `boolean`
>

### Type of change

- [x] Bug Fix (non-breaking change which fixes an issue)

### What problem does this PR solve?

The knowledge_graph chunk method is deprecated and should no longer be
used. #7184.

### Type of change

- [x] Refactoring

### What problem does this PR solve?

Enhance capability of `list_kbs`.

Breaking change: change method from `GET` to `POST`.

### Type of change

- [x] Refactoring
- [x] Enhancement with breaking change

### What problem does this PR solve?

https://github.com/infiniflow/ragflow/issues/7083

Internal due to when returning from ES, fields changed to str, so the
bool conversion does not work as expected.

### Type of change

- [x] Bug Fix (non-breaking change which fixes an issue)

### Related Issue:
https://github.com/infiniflow/ragflow/issues/6548

### Related PR:
https://github.com/infiniflow/ragflow/pull/6861


### Environment:
Commit version:
[[48730e0](48730e00a8)]

### Bug Description:
Unexpected `pymysql.err.InterfaceError: (0, '') `when using Peewee +
PyMySQL + PooledMySQLDatabase after a long-running `chat streamly`
operation.

This is a common issue with Peewee + PyMySQL + connection pooling: you
end up using a connection that was silently closed by the server, but
Peewee doesn't realize it's dead.

**I found that the error only occurs during longer streaming outputs**
and is unrelated to the database connection context, so it's likely
because:

- The prolonged streaming response caused the database connection to
time out

- The original database connection might have been disconnected by the
server during the streaming process

### Why This Happens
This error happens even when using `@DB.connection_context() `after the
stream is done. After investigation, I found this is caused by MySQL
connection pools that appear to be open but are actually dead (expired
due to` wait_timeout`).

1. `@DB.connection_context()` (as a decorator or context manager) pulls
a connection from the pool.

2. If this connection was idle and expired on the MySQL server (e.g.,
due to `wait_timeout`), but not closed in Python, it will still be
considered “open” (`DB.is_closed() == False`).

3. The real error will occur only when I execute a SQL command (such as
.`get_or_none()`), and PyMySQL tries to send it to the server via a
broken socket.


### Changes Made:

1. I implemented manual connection checks before executing SQL:
```
    try:
        DB.execute_sql("SELECT 1")
    except Exception:
        print("Connection dead, reconnecting...")
        DB.close()
        DB.connect()
```
2. Delayed the token count update until after the streaming response is
completed to ensure the streaming output isn't interrupted by database
operations.
```
        total_tokens = 0 
        for txt in chat_streamly(system, history, gen_conf):
            if isinstance(txt, int):
                total_tokens = txt
......
                break
......
        if total_tokens > 0:
            if not TenantLLMService.increase_usage(self.tenant_id, self.llm_type, txt, self.llm_name):
                logging.error("LLMBundle.chat_streamly can't update token usage for {}/CHAT llm_name: {}, content: {}".format(self.tenant_id, self.llm_name, txt))
```

### What problem does this PR solve?

https://github.com/infiniflow/ragflow/issues/6905

When deleting a document will check before removing it from storage

### Type of change

- [x] Bug Fix (non-breaking change which fixes an issue)

### What problem does this PR solve?


### Type of change

- [x] Refactoring