Add dify dsl demo (#3)

* Add dify dsl demo

* Deploying on docker

---------

Co-authored-by: Gerry.qi 齐晓明 <Gerry.qi@pousheng.com>
Co-authored-by: Gerry.qi <Gerry.qi@outlook.com>
This commit is contained in:
Gerry Qi
2025-06-27 16:28:58 +08:00
committed by GitHub
parent 124dd0da88
commit f135d9b949
3 changed files with 139 additions and 1 deletions

View File

@@ -222,9 +222,17 @@ Execute the following command to start the server:
```bash ```bash
./start_server.sh ./start_server.sh
``` ```
This command starts the FastAPI application with Streamable HTTP MCP service. This command starts the FastAPI application with Streamable HTTP MCP service.
### 5. Deploying on docker
If you want to run only Doris MCP Server in docker:
```bash
cd doris-mcp-server
docker build -t doris-mcp-server .
docker run -d -p <port>:<port> -v /*your-host*/doris-mcp-server/.env:/app/.env --name <your-mcp-server-name> -it doris-mcp-server:latest
```
**Service Endpoints:** **Service Endpoints:**
* **Streamable HTTP**: `http://<host>:<port>/mcp` (Primary MCP endpoint - supports GET, POST, DELETE, OPTIONS) * **Streamable HTTP**: `http://<host>:<port>/mcp` (Primary MCP endpoint - supports GET, POST, DELETE, OPTIONS)

View File

@@ -103,6 +103,9 @@ If your Dify deployment requires a publicly accessible endpoint, you can use the
2. Select **Agent** as the template and set the **App Name** (e.g., `Doris ChatBI`). 2. Select **Agent** as the template and set the **App Name** (e.g., `Doris ChatBI`).
![Agent setup](../images/dify_agent_setup.png) ![Agent setup](../images/dify_agent_setup.png)
3. Import from DSL,[dify_doris_dsl.yml](dify_doris_dsl.yml)
----- -----
## Instructions & Tool Configuration ## Instructions & Tool Configuration

View File

@@ -0,0 +1,127 @@
app:
description: ''
icon: 🤖
icon_background: '#FFEAD5'
mode: agent-chat
name: doris
use_icon_as_answer_icon: false
dependencies:
- current_identifier: null
type: marketplace
value:
marketplace_plugin_unique_identifier: langgenius/deepseek:0.0.5@21408d5c48cd9f18d66b08883d0999fe89e6d049c891324c2229dea23b9665d5
- current_identifier: null
type: marketplace
value:
marketplace_plugin_unique_identifier: junjiem/mcp_sse:0.2.1@53cc613667fcf91dd7208dd5f6d2c8df3c7ff0af8b79e8f3c0a430f1b39bda4c
kind: app
model_config:
agent_mode:
enabled: true
max_iteration: 10
prompt: null
strategy: function_call
tools:
- enabled: true
isDeleted: false
notAuthor: false
provider_id: junjiem/mcp_sse/mcp_sse
provider_name: junjiem/mcp_sse/mcp_sse
provider_type: builtin
tool_label: 获取 MCP 工具列表
tool_name: mcp_sse_list_tools
tool_parameters:
prompts_as_tools: 1
resources_as_tools: 1
servers_config: null
- enabled: true
isDeleted: false
notAuthor: false
provider_id: junjiem/mcp_sse/mcp_sse
provider_name: junjiem/mcp_sse/mcp_sse
provider_type: builtin
tool_label: 调用 MCP 工具
tool_name: mcp_sse_call_tool
tool_parameters:
arguments: ''
prompts_as_tools: ''
resources_as_tools: ''
servers_config: ''
tool_name: ''
annotation_reply:
enabled: false
chat_prompt_config: {}
completion_prompt_config: {}
dataset_configs:
datasets:
datasets: []
reranking_enable: true
reranking_mode: reranking_model
reranking_model:
reranking_model_name: ''
reranking_provider_name: ''
retrieval_model: multiple
top_k: 4
dataset_query_variable: ''
external_data_tools: []
file_upload:
allowed_file_extensions:
- .JPG
- .JPEG
- .PNG
- .GIF
- .WEBP
- .SVG
- .MP4
- .MOV
- .MPEG
- .WEBM
allowed_file_types: []
allowed_file_upload_methods:
- remote_url
- local_file
enabled: false
image:
detail: high
enabled: false
number_limits: 3
transfer_methods:
- remote_url
- local_file
number_limits: 3
model:
completion_params:
stop: []
mode: chat
name: deepseek-chat
provider: langgenius/deepseek/deepseek
more_like_this:
enabled: false
opening_statement: ''
pre_prompt: "<instruction>\nUse MCP tools to complete tasks as much as possible.\
\ Carefully read the annotations, method names, and parameter descriptions of\
\ each tool. Please follow these steps:\n1. Analyze the user's question and match\
\ the most appropriate tool.\n2. Use tool names and parameters exactly as defined;\
\ do not invent new ones.\n3. Pass parameters in the required JSON format.\n4.\
\ When calling tools, use:\n {\"mcp_sse_call_tool\": {\"tool_name\": \"<tool_name>\"\
, \"arguments\": \"{}\"}}\n5. Output plain text only—no XML tags.\n<input>\nUser\
\ question: user_query\n</input>\n<output>\nReturn tool results or a final answer,\
\ including analysis.\n</output>\n</instruction>"
prompt_type: simple
retriever_resource:
enabled: true
sensitive_word_avoidance:
configs: []
enabled: false
type: ''
speech_to_text:
enabled: false
suggested_questions: []
suggested_questions_after_answer:
enabled: false
text_to_speech:
enabled: false
language: ''
voice: ''
user_input_form: []
version: 0.3.0