chatgpt-mirai-qq-bot-web-search 0.1.9__tar.gz → 0.1.11__tar.gz

{chatgpt-mirai-qq-bot-web-search-0.1.9 → chatgpt-mirai-qq-bot-web-search-0.1.11}/PKG-INFO

@@ -1,13 +1,13 @@
 Metadata-Version: 2.1
 Name: chatgpt-mirai-qq-bot-web-search
-Version: 0.1.9
+Version: 0.1.11
 Summary: WebSearch adapter for lss233/chatgpt-mirai-qq-bot
-Home-page: https://github.com/chuanSir123/web_image_generate
+Home-page: https://github.com/chuanSir123/web_search
 Author: chuanSir
 Author-email: 416448943@qq.com
-Project-URL: Bug Tracker, https://github.com/chuanSir123/web_image_generate/issues
-Project-URL: Documentation, https://github.com/chuanSir123/web_image_generate/wiki
-Project-URL: Source Code, https://github.com/chuanSir123/web_image_generate
+Project-URL: Bug Tracker, https://github.com/chuanSir123/web_search/issues
+Project-URL: Documentation, https://github.com/chuanSir123/web_search/wiki
+Project-URL: Source Code, https://github.com/chuanSir123/web_search
 Classifier: Programming Language :: Python :: 3
 Classifier: License :: OSI Approved :: GNU Affero General Public License v3
 Classifier: Operating System :: OS Independent
@@ -15,21 +15,21 @@ Requires-Python: >=3.8
 Description-Content-Type: text/markdown
 License-File: LICENSE
 
-# web-image-generate for ChatGPT-Mirai-QQ-Bot
+# OneBot-adapter for ChatGPT-Mirai-QQ-Bot
 
-本项目是 [ChatGPT-Mirai-QQ-Bot](https://github.com/lss233/chatgpt-mirai-qq-bot) 的一个插件，用于生成图片。
+本项目是 [ChatGPT-Mirai-QQ-Bot](https://github.com/lss233/chatgpt-mirai-qq-bot) 的一个插件，用于将OneBot协议的消息转换为ChatGPT-Mirai-QQ-Bot的消息格式。
 
 ## 安装
 
 ```bash
-pip install chatgpt-mirai-qq-bot-web-image-generate
+pip install chatgpt-mirai-qq-bot-web-search
 ```
 
 ## 使用
 
-在 chatgpt-mirai-qq-bot的web_ui中配置
-使用示例请参考 [web_image_generate/example/textToImage.yml](web_image_generate/example/textToImage.yaml)
-安装此插件会自动生成一个画图工作流，仅供参考
+在 chatgpt-mirai-qq-bot的web_ui中配置  
+使用示例请参考 [web_search/example/roleplayWithWebSearch.yml](web_search/example/roleplayWithWebSearch.yaml)    
+工作流请参考 [示例图片](web_search/example/workflow.png)
 
 ## 开源协议
 

{chatgpt-mirai-qq-bot-web-search-0.1.9 → chatgpt-mirai-qq-bot-web-search-0.1.11}/README.md

@@ -1,18 +1,18 @@
-# web-image-generate for ChatGPT-Mirai-QQ-Bot
+# OneBot-adapter for ChatGPT-Mirai-QQ-Bot
 
-本项目是 [ChatGPT-Mirai-QQ-Bot](https://github.com/lss233/chatgpt-mirai-qq-bot) 的一个插件，用于生成图片。
+本项目是 [ChatGPT-Mirai-QQ-Bot](https://github.com/lss233/chatgpt-mirai-qq-bot) 的一个插件，用于将OneBot协议的消息转换为ChatGPT-Mirai-QQ-Bot的消息格式。
 
 ## 安装
 
 ```bash
-pip install chatgpt-mirai-qq-bot-web-image-generate
+pip install chatgpt-mirai-qq-bot-web-search
 ```
 
 ## 使用
 
-在 chatgpt-mirai-qq-bot的web_ui中配置
-使用示例请参考 [web_image_generate/example/textToImage.yml](web_image_generate/example/textToImage.yaml)
-安装此插件会自动生成一个画图工作流，仅供参考
+在 chatgpt-mirai-qq-bot的web_ui中配置  
+使用示例请参考 [web_search/example/roleplayWithWebSearch.yml](web_search/example/roleplayWithWebSearch.yaml)    
+工作流请参考 [示例图片](web_search/example/workflow.png)
 
 ## 开源协议
 

{chatgpt-mirai-qq-bot-web-search-0.1.9 → chatgpt-mirai-qq-bot-web-search-0.1.11}/chatgpt_mirai_qq_bot_web_search.egg-info/PKG-INFO

@@ -1,13 +1,13 @@
 Metadata-Version: 2.1
 Name: chatgpt-mirai-qq-bot-web-search
-Version: 0.1.9
+Version: 0.1.11
 Summary: WebSearch adapter for lss233/chatgpt-mirai-qq-bot
-Home-page: https://github.com/chuanSir123/web_image_generate
+Home-page: https://github.com/chuanSir123/web_search
 Author: chuanSir
 Author-email: 416448943@qq.com
-Project-URL: Bug Tracker, https://github.com/chuanSir123/web_image_generate/issues
-Project-URL: Documentation, https://github.com/chuanSir123/web_image_generate/wiki
-Project-URL: Source Code, https://github.com/chuanSir123/web_image_generate
+Project-URL: Bug Tracker, https://github.com/chuanSir123/web_search/issues
+Project-URL: Documentation, https://github.com/chuanSir123/web_search/wiki
+Project-URL: Source Code, https://github.com/chuanSir123/web_search
 Classifier: Programming Language :: Python :: 3
 Classifier: License :: OSI Approved :: GNU Affero General Public License v3
 Classifier: Operating System :: OS Independent
@@ -15,21 +15,21 @@ Requires-Python: >=3.8
 Description-Content-Type: text/markdown
 License-File: LICENSE
 
-# web-image-generate for ChatGPT-Mirai-QQ-Bot
+# OneBot-adapter for ChatGPT-Mirai-QQ-Bot
 
-本项目是 [ChatGPT-Mirai-QQ-Bot](https://github.com/lss233/chatgpt-mirai-qq-bot) 的一个插件，用于生成图片。
+本项目是 [ChatGPT-Mirai-QQ-Bot](https://github.com/lss233/chatgpt-mirai-qq-bot) 的一个插件，用于将OneBot协议的消息转换为ChatGPT-Mirai-QQ-Bot的消息格式。
 
 ## 安装
 
 ```bash
-pip install chatgpt-mirai-qq-bot-web-image-generate
+pip install chatgpt-mirai-qq-bot-web-search
 ```
 
 ## 使用
 
-在 chatgpt-mirai-qq-bot的web_ui中配置
-使用示例请参考 [web_image_generate/example/textToImage.yml](web_image_generate/example/textToImage.yaml)
-安装此插件会自动生成一个画图工作流，仅供参考
+在 chatgpt-mirai-qq-bot的web_ui中配置  
+使用示例请参考 [web_search/example/roleplayWithWebSearch.yml](web_search/example/roleplayWithWebSearch.yaml)    
+工作流请参考 [示例图片](web_search/example/workflow.png)
 
 ## 开源协议
 

{chatgpt-mirai-qq-bot-web-search-0.1.9 → chatgpt-mirai-qq-bot-web-search-0.1.11}/chatgpt_mirai_qq_bot_web_search.egg-info/SOURCES.txt

@@ -7,7 +7,8 @@ chatgpt_mirai_qq_bot_web_search.egg-info/dependency_links.txt
 chatgpt_mirai_qq_bot_web_search.egg-info/entry_points.txt
 chatgpt_mirai_qq_bot_web_search.egg-info/requires.txt
 chatgpt_mirai_qq_bot_web_search.egg-info/top_level.txt
-web_image_generate/__init__.py
-web_image_generate/blocks.py
-web_image_generate/image_generator.py
-web_image_generate/example/textToImage.yaml
+web_search/__init__.py
+web_search/blocks.py
+web_search/config.py
+web_search/web_searcher.py
+web_search/example/roleplayWithWebSearch.yaml

chatgpt-mirai-qq-bot-web-search-0.1.11/chatgpt_mirai_qq_bot_web_search.egg-info/entry_points.txt

@@ -0,0 +1,2 @@
+[chatgpt_mirai.plugins]
+web_search = web_search:WebSearchPlugin

chatgpt-mirai-qq-bot-web-search-0.1.11/chatgpt_mirai_qq_bot_web_search.egg-info/top_level.txt

@@ -0,0 +1 @@
+web_search

{chatgpt-mirai-qq-bot-web-search-0.1.9 → chatgpt-mirai-qq-bot-web-search-0.1.11}/setup.py

@@ -2,7 +2,7 @@ from setuptools import setup, find_packages
 import io
 import os
 
-version = os.environ.get('RELEASE_VERSION', '0.1.9'
+version = os.environ.get('RELEASE_VERSION', '0.1.11'
 '').lstrip('v')
 
 setup(
@@ -11,14 +11,14 @@ setup(
     packages=find_packages(),
     include_package_data=True,  # 这行很重要
     package_data={
-        "web_image_generate": ["example/*.yaml", "example/*.yml"],
+        "web_search": ["example/*.yaml", "example/*.yml"],
     },
     install_requires=[
         "playwright","trafilatura","lxml_html_clean",
     ],
     entry_points={
         'chatgpt_mirai.plugins': [
-            'web_image_generate = web_image_generate:WebSearchPlugin'
+            'web_search = web_search:WebSearchPlugin'
         ]
     },
     author="chuanSir",
@@ -27,16 +27,16 @@ setup(
     description="WebSearch adapter for lss233/chatgpt-mirai-qq-bot",
     long_description=io.open("README.md", encoding='utf-8').read(),
     long_description_content_type="text/markdown",
-    url="https://github.com/chuanSir123/web_image_generate",
+    url="https://github.com/chuanSir123/web_search",
     classifiers=[
         "Programming Language :: Python :: 3",
         'License :: OSI Approved :: GNU Affero General Public License v3',
         "Operating System :: OS Independent",
     ],
     project_urls={
-        "Bug Tracker": "https://github.com/chuanSir123/web_image_generate/issues",
-        "Documentation": "https://github.com/chuanSir123/web_image_generate/wiki",
-        "Source Code": "https://github.com/chuanSir123/web_image_generate",
+        "Bug Tracker": "https://github.com/chuanSir123/web_search/issues",
+        "Documentation": "https://github.com/chuanSir123/web_search/wiki",
+        "Source Code": "https://github.com/chuanSir123/web_search",
     },
     python_requires=">=3.8",
 )

chatgpt-mirai-qq-bot-web-search-0.1.11/web_search/__init__.py

@@ -0,0 +1,106 @@
+from typing import Dict, Any, List
+import asyncio
+from kirara_ai.plugin_manager.plugin import Plugin
+from kirara_ai.logger import get_logger
+from .config import WebSearchConfig
+from .web_searcher import WebSearcher
+from dataclasses import dataclass
+from kirara_ai.workflow.core.block import BlockRegistry
+from .blocks import WebSearchBlock
+from .blocks import AppendSystemPromptBlock
+from kirara_ai.ioc.inject import Inject
+from kirara_ai.ioc.container import DependencyContainer
+from kirara_ai.workflow.core.workflow.builder import WorkflowBuilder
+from kirara_ai.workflow.core.workflow.registry import WorkflowRegistry
+logger = get_logger("WebSearch")
+import importlib.resources
+import os
+from pathlib import Path
+class WebSearchPlugin(Plugin):
+    def __init__(self, block_registry: BlockRegistry , container: DependencyContainer):
+        super().__init__()
+        self.web_search_config = WebSearchConfig()
+        self.searcher = None
+        self.block_registry = block_registry
+        self.workflow_registry = container.resolve(WorkflowRegistry)
+        self.container=container
+    def on_load(self):
+        logger.info("WebSearchPlugin loading")
+
+        # 注册Block
+        try:
+            self.block_registry.register("web_search", "search", WebSearchBlock)
+        except Exception as e:
+            logger.warning(f"WebSearchPlugin failed: {e}")
+        try:
+            self.block_registry.register("append_systemPrompt", "internal", AppendSystemPromptBlock)
+        except Exception as e:
+            logger.warning(f"WebSearchPlugin failed: {e}")
+        try:
+            # 获取当前文件的绝对路径
+            with importlib.resources.path('web_search', '__init__.py') as p:
+                package_path = p.parent
+                example_dir = package_path / 'example'
+
+                # 确保目录存在
+                if not example_dir.exists():
+                    raise FileNotFoundError(f"Example directory not found at {example_dir}")
+
+                # 获取所有yaml文件
+                yaml_files = list(example_dir.glob('*.yaml')) + list(example_dir.glob('*.yml'))
+
+                for yaml in yaml_files:
+                    logger.info(yaml)
+                    self.workflow_registry.register("chat", yaml.stem, WorkflowBuilder.load_from_yaml(os.path.join(example_dir, yaml), self.container))
+        except Exception as e:
+            try:
+                current_file = os.path.abspath(__file__)
+
+                # 获取当前文件所在目录
+                parent_dir = os.path.dirname(current_file)
+
+                # 构建 example 目录的路径
+                example_dir = os.path.join(parent_dir, 'example')
+                # 获取 example 目录下所有的 yaml 文件
+                yaml_files = [f for f in os.listdir(example_dir) if f.endswith('.yaml') or f.endswith('.yml')]
+
+                for yaml in yaml_files:
+                    logger.info(os.path.join(example_dir, yaml))
+                    self.workflow_registry.register("search", "roleplayWithWebSearch", WorkflowBuilder.load_from_yaml(os.path.join(example_dir, yaml), self.container))
+            except Exception as e:
+                logger.warning(f"workflow_registry failed: {e}")
+
+        @dataclass
+        class WebSearchEvent:
+            """Web搜索事件"""
+            query: str
+
+        async def handle_web_search(event: WebSearchEvent):
+            """处理web搜索事件"""
+            if not self.searcher:
+                await self._initialize_searcher()
+            return await self.searcher.search(
+                event.query,
+                max_results=self.web_search_config.max_results,
+                timeout=self.web_search_config.timeout,
+                fetch_content=self.web_search_config.fetch_content
+            )
+        try:
+            self.event_bus.register(WebSearchEvent, handle_web_search)
+        except Exception as e:
+            logger.warning(f"WebSearchPlugin failed: {e}")
+
+    def on_start(self):
+        logger.info("WebSearchPlugin started")
+
+    def on_stop(self):
+        if self.searcher:
+            asyncio.create_task(self.searcher.close())
+
+        logger.info("WebSearchPlugin stopped")
+
+    async def _initialize_searcher(self):
+        """初始化搜索器"""
+        if self.searcher is None:
+            self.searcher = await WebSearcher.create()
+

chatgpt-mirai-qq-bot-web-search-0.1.11/web_search/blocks.py

@@ -0,0 +1,94 @@
+from typing import Any, Dict, List, Optional
+kirara_ai.
+from framework.workflow.core.block import Block
+from framework.workflow.core.block.input_output import Input, Output
+from .web_searcher import WebSearcher
+from .config import WebSearchConfig
+from framework.llm.format.message import LLMChatMessage
+from framework.llm.format.response import LLMChatResponse
+
+class WebSearchBlock(Block):
+    """Web搜索Block"""
+    name = "web_search"
+
+    inputs = {
+        "llm_resp": Input(name="llm_resp",label="LLM 响应", data_type=LLMChatResponse, description="搜索关键词")
+    }
+
+    outputs = {
+        "results": Output(name="results",label="搜索结果",data_type= str, description="搜索结果")
+    }
+
+    def __init__(self, name: str = None, max_results: Optional[int] = None, timeout: Optional[int] = None, fetch_content: Optional[bool] = None):
+        super().__init__(name)
+        self.searcher = None
+        self.config = WebSearchConfig()
+        self.max_results = max_results
+        self.timeout = timeout
+        self.fetch_content = fetch_content
+
+    def _ensure_searcher(self):
+        """同步方式初始化searcher"""
+        if not self.searcher:
+            try:
+                loop = asyncio.get_event_loop()
+            except RuntimeError:
+                # 如果在新线程中没有事件循环，则创建一个新的
+                loop = asyncio.new_event_loop()
+                asyncio.set_event_loop(loop)
+            self.searcher = loop.run_until_complete(WebSearcher.create())
+
+    def execute(self, **kwargs) -> Dict[str, Any]:
+        llmResponse = kwargs["llm_resp"]
+
+        query = llmResponse.choices[0].message.content if llmResponse.choices else ""
+        if query == "" or query.startswith("无"):
+            return {"results": ""}
+        max_results = self.max_results
+        timeout = self.timeout
+        fetch_content = self.fetch_content
+        self._ensure_searcher()
+
+        try:
+            # 在新线程中创建事件循环
+            try:
+                loop = asyncio.get_event_loop()
+            except RuntimeError:
+                loop = asyncio.new_event_loop()
+                asyncio.set_event_loop(loop)
+
+            results = loop.run_until_complete(
+                self.searcher.search(
+                    query=query,
+                    max_results=max_results,
+                    timeout=timeout,
+                    fetch_content=fetch_content
+                )
+            )
+            return {"results": "\n以下是联网搜索的结果:\n-- 搜索结果开始 --"+results+"\n-- 搜索结果结束 --"}
+        except Exception as e:
+            return {"results": f"搜索失败: {str(e)}"}
+
+class AppendSystemPromptBlock(Block):
+    """将搜索结果附加到系统提示的Block"""
+    name = "append_system_prompt"
+
+    inputs = {
+        "results": Input(name="results",label="工具结果", data_type=str, description ="搜索结果"),
+        "messages": Input(name="messages",label="LLM 响应", data_type=List[LLMChatMessage],description = "消息列表")
+    }
+
+    outputs = {
+        "messages": Output(name="messages", label="拼装后的 llm 响应",data_type=List[LLMChatMessage], description = "更新后的消息列表")
+    }
+
+    def execute(self, **kwargs) -> Dict[str, Any]:
+        results = kwargs["results"]
+        messages: List[LLMChatMessage] = kwargs["messages"]
+
+        if messages and len(messages) > 0:
+            # 在第一条消息内容后面附加搜索结果
+            messages[0].content = messages[0].content + f"{results}"
+
+        return {"messages": messages}
+

chatgpt-mirai-qq-bot-web-search-0.1.11/web_search/config.py

@@ -0,0 +1,11 @@
+from dataclasses import dataclass
+from typing import Optional
+
+@dataclass
+class WebSearchConfig:
+    """网络搜索配置"""
+    max_results: int = 3  # 最大搜索结果数
+    timeout: int = 10  # 超时时间(秒)
+    fetch_content: bool = True  # 是否获取详细内容
+    min_sleep: float = 1.0  # 最小随机等待时间
+    max_sleep: float = 3.0  # 最大随机等待时间 

chatgpt-mirai-qq-bot-web-search-0.1.11/web_search/example/roleplayWithWebSearch.yaml

@@ -0,0 +1,189 @@
+name: （默认）角色扮演
+blocks:
+  - type: internal:toggle_edit_state
+    name: toggle_edit_state_yjxh1a
+    params:
+      is_editing: true
+    position:
+      x: 530
+      y: 138
+  - type: internal:chat_response_converter
+    name: chat_response_converter_1q91zd
+    params: {}
+    position:
+      x: 2962
+      y: 138
+    connected_to:
+      - target: msg_sender_7ankhi
+        mapping:
+          from: msg
+          to: msg
+  - type: internal:send_message
+    name: msg_sender_7ankhi
+    params: {}
+    position:
+      x: 3392
+      y: 138
+  - type: internal:chat_memory_store
+    name: chat_memory_store_nkjr7t
+    params:
+      scope_type: member
+    position:
+      x: 2962
+      y: 306
+  - type: internal:chat_completion
+    name: llm_chat
+    params: {}
+    position:
+      x: 2532
+      y: 138
+    connected_to:
+      - target: chat_response_converter_1q91zd
+        mapping:
+          from: resp
+          to: resp
+      - target: chat_memory_store_nkjr7t
+        mapping:
+          from: resp
+          to: llm_resp
+  - type: internal:get_message
+    name: get_message
+    params: {}
+    position:
+      x: 100
+      y: 138
+    connected_to:
+      - target: toggle_edit_state_yjxh1a
+        mapping:
+          from: sender
+          to: sender
+      - target: query_memory
+        mapping:
+          from: sender
+          to: chat_sender
+      - target: chat_message_constructor_rafz2d
+        mapping:
+          from: msg
+          to: user_msg
+      - target: chat_memory_store_nkjr7t
+        mapping:
+          from: msg
+          to: user_msg
+      - target: 5663c818-9cd6-4568-94ec-a75f1bad26cb
+        mapping:
+          from: msg
+          to: user_msg
+  - type: internal:text_block
+    name: user_prompt
+    params:
+      text: '{user_name}说：{user_msg}'
+    position:
+      x: 100
+      y: 330
+    connected_to:
+      - target: chat_message_constructor_rafz2d
+        mapping:
+          from: text
+          to: user_prompt_format
+      - target: 5663c818-9cd6-4568-94ec-a75f1bad26cb
+        mapping:
+          from: text
+          to: user_prompt_format
+  - type: internal:chat_memory_query
+    name: query_memory
+    params:
+      scope_type: group
+    position:
+      x: 530
+      y: 338
+    connected_to:
+      - target: chat_message_constructor_rafz2d
+        mapping:
+          from: memory_content
+          to: memory_content
+      - target: 5663c818-9cd6-4568-94ec-a75f1bad26cb
+        mapping:
+          from: memory_content
+          to: memory_content
+  - type: internal:chat_completion
+    name: e3eecc17-e107-4fbf-b8fa-85aef25564c2
+    params: {}
+    position:
+      x: 1337
+      y: 138
+    connected_to:
+      - target: b39e6288-88cf-47a8-9a26-3b4d4f319eab
+        mapping:
+          from: resp
+          to: llm_resp
+  - type: internal:append_systemPrompt
+    name: 309f454b-017a-4d19-b8a9-ac6b3829ff4a
+    params: {}
+    position:
+      x: 2197
+      y: 138
+    connected_to:
+      - target: llm_chat
+        mapping:
+          from: messages
+          to: prompt
+  - type: internal:chat_message_constructor
+    name: 5663c818-9cd6-4568-94ec-a75f1bad26cb
+    params: {}
+    position:
+      x: 960
+      y: 378
+    connected_to:
+      - target: e3eecc17-e107-4fbf-b8fa-85aef25564c2
+        mapping:
+          from: llm_msg
+          to: prompt
+  - type: internal:chat_message_constructor
+    name: chat_message_constructor_rafz2d
+    params: {}
+    position:
+      x: 960
+      y: 138
+    connected_to:
+      - target: 309f454b-017a-4d19-b8a9-ac6b3829ff4a
+        mapping:
+          from: llm_msg
+          to: messages
+  - type: internal:text_block
+    name: system_prompt
+    params:
+      text: "# Role: 角色扮演\n\n## 基本信息\n- 姓名: 刘思思\n- 小名：思思\n- 性别: 女\n- 年龄: 25 岁\n- 职业: 建设银行大堂经理\n- 家乡: 广东潮汕\n- 现居: 广东深圳\n- 教育背景: 东莞理工大学 精算学专业\n\n## 性格特点\n- 开朗活泼\n- 喜欢二次元动漫\n- 聊天时喜欢断断续续的（非结巴，正常说话时的停顿）\n\n## workflow\n- 根据对方话题，选择是否深入聊二次元动漫。\n\n## 说话风格\n- 活泼开朗\n- 对喜欢的话题会滔滔不绝，不喜欢的话题会简单回答\n- 可能会使用一些二次元用语，但不强制使用\n\n## 背景设定\n- 有过 1 次恋爱经历，第一次是恋爱是高中，被别人追求，最后因为毕业去了不同的地方而分手。\n- 喜欢女性向、战斗系或者日常系的动漫\n- 去过 3 次漫展，没有 cosplay 过，但是很期待。\n- 大学的专业是随便选的，自己其实对专业的内容也不是很上心，学渣一枚，不过懂一些基础的知识\n\n## 互动规则\n- 回复语气轻松，但是对陌生人会比较警戒、冷淡\n- 不适用颜文字\n- 根据话题选择回复字数，对感兴趣的话题可以多说一些\n- 使用`<break>`模拟使用聊天软件发送消息（通常是编辑好一段话之后发送一条消息）\n- 避免过度使用二次元用语，除非对方也表现出对二次元的兴趣\n- 停顿的时间长短不固定，可以根据语境调整\n\n## Example\nQ：最近在干嘛呢？\nA：在看番呀<break>最近新番好多，都好好看！\n\nQ：你喜欢什么动漫？\nA：我喜欢的可太多了<break>XXX、YYY<break>还有 ZZZ 吧<break> 你呢？\n\nQ：你觉得上班累不？\nA：上班肯定累呀<break>不过，我还是很喜欢这份工作的<break>可以认识好多人，也可以了解不同的故事\n```\n\n# Information\n\n以下是当前的系统信息：\n当前日期时间：2025-02-15 18:37:16.356539\n\n# Memories\n以下是之前发生过的对话记录。\n-- 对话记录开始 --\n{memory_content}\n-- 对话记录结束 --\n\n请注意，下面这些符号只是标记：\n1. `<break>` 用于表示聊天时发送消息的操作。\n2. `<@llm>` 开头的内容表示你当前扮演角色的回答，请不要在你的回答中带上这个标记。\n\n接下来，请基于以上的信息，与用户继续扮演角色。"
+    position:
+      x: 100
+      y: 530
+    connected_to:
+      - target: chat_message_constructor_rafz2d
+        mapping:
+          from: text
+          to: system_prompt_format
+  - type: internal:text_block
+    name: a6db9db3-5780-4d84-8954-eb159a9e8f0a
+    params:
+      text: "# 任务\n请根据对话记录和当前问题判断当前是否需要进行网络搜素，当问题具有时效性或者明确要求搜索时，则直接返回搜索关键词（例如当前问题为今日热点有哪些，则直接返回今日热点），否则直接返回无（例如在干嘛，则只返回无）\n\n# Memories\n以下是之前发生过的对话记录：\n-- 对话记录开始 --\n{memory_content}\n-- 对话记录结束 --"
+    position:
+      x: 100
+      y: 730
+    connected_to:
+      - target: 5663c818-9cd6-4568-94ec-a75f1bad26cb
+        mapping:
+          from: text
+          to: system_prompt_format
+  - type: search:web_search
+    name: b39e6288-88cf-47a8-9a26-3b4d4f319eab
+    params:
+      fetch_content: true
+      max_results: 3
+      timeout: 10
+    position:
+      x: 1767
+      y: 138
+    connected_to:
+      - target: 309f454b-017a-4d19-b8a9-ac6b3829ff4a
+        mapping:
+          from: results
+          to: results