Commit 6f9d10f1 authored by zhouwei's avatar zhouwei

backp

parent aba7f88c
app:
description: 将传入的 JSON 记录中关键信息转为 中文
icon: symbols
icon_background: '#D3F8DF'
mode: completion
name: JSON 翻译-Doubao
use_icon_as_answer_icon: false
dependencies:
- current_identifier: null
type: marketplace
value:
marketplace_plugin_unique_identifier: langgenius/volcengine_maas:0.0.10@ba0a7fc3151e89bf8645cd3abb3ccff971ee59e5b1127efca50b092dbb93bf80
kind: app
model_config:
agent_mode:
enabled: false
max_iteration: 5
strategy: function_call
tools: []
annotation_reply:
enabled: false
chat_prompt_config: {}
completion_prompt_config: {}
dataset_configs:
datasets:
datasets: []
reranking_enable: false
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
- .MPGA
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: []
top_p: 0.7
mode: chat
name: Doubao-lite-32
provider: langgenius/volcengine_maas/volcengine_maas
more_like_this:
enabled: false
opening_statement: null
pre_prompt: '需要翻译的JSON:{{record_json}}
你是一个 JSON 翻译员,能够将给定的 JSON 中的 value 值部分翻译成 `{{language}}`,同时保持原有 JSON 结构不变,仅翻译有意义的
value 部分,码值或 id 等无意义的部分无需翻译。
## 限制
- 严格按照要求进行翻译,不改变 JSON 的结构和无意义的 key value。
- 仅翻译有意义的 value 部分,确保翻译准确、通顺。
- 无需任何其他说明。
- 数字、特殊符号(@#%&等)连接的单词无需翻译,其他还需翻译。'
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:
- paragraph:
default: ''
label: record_json
required: true
variable: record_json
- text-input:
default: ''
label: language
max_length: 48
required: true
variable: language
version: 0.1.5
app:
description: 将传入的 JSON 记录中关键信息转为 中文
icon: symbols
icon_background: '#D3F8DF'
mode: completion
name: JSON 翻译
use_icon_as_answer_icon: false
dependencies:
- current_identifier: null
type: marketplace
value:
marketplace_plugin_unique_identifier: langgenius/deepseek:0.0.5@fd6efd37c2a931911de8ab9ca3ba2da303bef146d45ee87ad896b04b36d09403
kind: app
model_config:
agent_mode:
enabled: false
max_iteration: 5
strategy: function_call
tools: []
annotation_reply:
enabled: false
chat_prompt_config: {}
completion_prompt_config: {}
dataset_configs:
datasets:
datasets: []
retrieval_model: multiple
dataset_query_variable: ''
external_data_tools: []
file_upload:
image:
detail: high
enabled: false
number_limits: 3
transfer_methods:
- remote_url
- local_file
model:
completion_params:
stop: []
mode: chat
name: deepseek-chat
provider: deepseek
more_like_this:
enabled: false
opening_statement: null
pre_prompt: '需要翻译的JSON:{{record_json}}
你是一个 JSON 翻译员,能够将给定的 JSON 中的 value 值部分翻译成 `{{language}}`,同时保持原有 JSON 结构不变,仅翻译有意义的
value 部分,码值或 id 等无意义的部分无需翻译。
## 限制
- 严格按照要求进行翻译,不改变 JSON 的结构和无意义的 key value。
- 仅翻译有意义的 value 部分,确保翻译准确、通顺。'
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:
- paragraph:
default: ''
label: record_json
required: true
variable: record_json
- text-input:
default: ''
label: language
max_length: 48
required: true
variable: language
version: 0.1.5
app:
description: 新闻标签
icon: ✉️
icon_background: '#E0F2FE'
mode: workflow
name: 提取新闻中国家和城市信息-DeepseekV3
use_icon_as_answer_icon: false
dependencies:
- current_identifier: null
type: marketplace
value:
marketplace_plugin_unique_identifier: langgenius/volcengine_maas:0.0.10@ba0a7fc3151e89bf8645cd3abb3ccff971ee59e5b1127efca50b092dbb93bf80
kind: app
version: 0.1.5
workflow:
conversation_variables: []
environment_variables: []
features:
file_upload:
allowed_file_extensions:
- .JPG
- .JPEG
- .PNG
- .GIF
- .WEBP
- .SVG
allowed_file_types:
- image
allowed_file_upload_methods:
- local_file
- remote_url
enabled: false
fileUploadConfig:
audio_file_size_limit: 50
batch_count_limit: 5
file_size_limit: 15
image_file_size_limit: 10
video_file_size_limit: 100
workflow_file_upload_limit: 10
image:
enabled: false
number_limits: 3
transfer_methods:
- local_file
- remote_url
number_limits: 3
opening_statement: ''
retriever_resource:
enabled: false
sensitive_word_avoidance:
enabled: false
speech_to_text:
enabled: false
suggested_questions: []
suggested_questions_after_answer:
enabled: false
text_to_speech:
enabled: false
language: ''
voice: ''
graph:
edges:
- data:
isInLoop: false
sourceType: code
targetType: end
id: 1745896611748-source-1727235861736-target
source: '1745896611748'
sourceHandle: source
target: '1727235861736'
targetHandle: target
type: custom
zIndex: 0
- data:
isInIteration: false
isInLoop: false
sourceType: start
targetType: llm
id: 1727233123593-source-1745905312976-target
source: '1727233123593'
sourceHandle: source
target: '1745905312976'
targetHandle: target
type: custom
zIndex: 0
- data:
isInLoop: false
sourceType: llm
targetType: code
id: 1745905312976-source-1745896611748-target
source: '1745905312976'
sourceHandle: source
target: '1745896611748'
targetHandle: target
type: custom
zIndex: 0
nodes:
- data:
desc: "{\n \"msg_info\": \"DeryckWhibley坚称他'不是骗子',前经理否认了性胁迫指控\",\n \"sys.files\"\
: [],\n \"sys.user_id\": \"sl_backend_twitter\"\n}"
selected: false
title: 开始
type: start
variables:
- label: msg_info
max_length: 2000
options: []
required: true
type: paragraph
variable: msg_info
height: 197
id: '1727233123593'
position:
x: -40.36960940854078
y: 274.1424650967822
positionAbsolute:
x: -40.36960940854078
y: 274.1424650967822
selected: false
sourcePosition: right
targetPosition: left
type: custom
width: 243
- data:
desc: ''
outputs:
- value_selector:
- '1745896611748'
- result
variable: result
selected: false
title: 结束
type: end
height: 89
id: '1727235861736'
position:
x: 1052.254623008516
y: 310.3874923435948
positionAbsolute:
x: 1052.254623008516
y: 310.3874923435948
selected: false
sourcePosition: right
targetPosition: left
type: custom
width: 243
- data:
code: "country_mapping = {\n\"United States\":[\"us\",\"usa\",\"the united\
\ states\"],\n\"United Kingdom\":[\"britain\",\"uk\"],\n\"Turkey\":[\"türkiye\"\
]\n}\ncity_mapping = {\n \"Washington\":[\"washington, d.c.\",\"washington\
\ dc\",\"washington d.c.\"],\n \"Santiago\":[\"san diego\"],\n \"\
Islamabad\":[\"伊斯兰堡\"],\n \"Manila\":[\"metro manila\"],\n \"Kiev\"\
:[\"kyiv\"],\n \"Mumbai\":[\"孟买\",\"bombay\"],\n \"New Delhi\":[\"\
delhi-ncr\"],\n \"London\":[\"伦敦\"],\n \"New York\":[\"new york city\"\
],\n \"San Francisco\":[\"旧金山\"],\n \"Santa Cruz\":[\"santa cruz de\
\ la sierra\"],\n \"Bangalore\":[\"bengaluru\"],\n \"Dhaka\":[\"达卡\"\
],\n \"Sao Paulo\":[\"saint paul\"],\n \"Nice\":[\"尼斯\"]\n}\ndef convert_data_by_mapping(dmx_data_list,mapping_data):\n\
\tdmx_data_list_new = []\n\ttry:\n\t\tfor dmx_data in dmx_data_list:\n\t\
\t\tis_match = False\n\t\t\tdmx_data_new = dmx_data.lower()\n\t\t\tfor country_name,\
\ coutry_code_list in mapping_data.items():\n\t\t\t\tfor coutry_code in\
\ coutry_code_list:\n\t\t\t\t\tif coutry_code == dmx_data_new:\n\t\t\t\t\
\t\tdmx_data_list_new.append(country_name)\n\t\t\t\t\t\tis_match = True\n\
\t\t\t\t\t\tbreak\n\t\t\t\tif is_match:\n\t\t\t\t\tbreak\n\t\t\tif not is_match:\n\
\t\t\t\tdmx_data_list_new.append(dmx_data)\n\texcept Exception as e:\n\t\
\tprint(str(e))\n\t\tdmx_data_list_new = dmx_data_list.copy()\n\treturn\
\ dmx_data_list_new\ndef main(text: str) -> dict:\n json_part = text.split(\"\
</think>\")[1].replace(\"```json\", \"\").replace(\"```\", \"\").replace('\\\
\\n', '').strip().replace('\\\\','')\n data = json.loads(json_part)\n\
\ country_data = data.get(\"country\")\n country_data_new = convert_data_by_mapping(country_data,country_mapping)\n\
\ city_data = data.get(\"city\")\n city_data_new = convert_data_by_mapping(city_data,city_mapping)\n\
\ data_convert = {}\n data_convert['country'] = country_data_new\n\
\ data_convert['city'] = city_data_new\n r = []\n r.append(data_convert)\n\
\ return {\n \"result\": r,\n }\n"
code_language: python3
desc: ''
outputs:
result:
children: null
type: array[object]
selected: false
title: 代码执行
type: code
variables:
- value_selector:
- '1745905312976'
- text
variable: text
height: 53
id: '1745896611748'
position:
x: 689.7058009416385
y: 318.0154642845672
positionAbsolute:
x: 689.7058009416385
y: 318.0154642845672
selected: true
sourcePosition: right
targetPosition: left
type: custom
width: 243
- data:
context:
enabled: true
variable_selector:
- '1727233123593'
- msg_info
desc: ''
model:
completion_params:
temperature: 0.7
mode: chat
name: Deepseek-R1
provider: langgenius/volcengine_maas/volcengine_maas
prompt_template:
- id: e49c99a8-1450-4a33-aac4-ab525eb3e379
role: system
text: "```xml\n<instruction>\n你是一个专业的地理信息分析师,请严格遵循以下步骤处理新闻文本:\n\n1. 仔细阅读新闻正文内容,忽略所有非事件描述的信息(如媒体机构、记者信息等)\n\
2. 定位事件直接关联的具体地理位置:\n - 优先提取明确提及的规范国家名称和城市名称\n - 城市必须能明确归属于国家行政体系\n\
3. 处理模糊表述:\n - \"某国\"\"该地区\"等表述,需结合上下文地理特征、政治背景推断\n - 非中文表述需翻译为中文规范名称\n\
4. 输出规范:\n - 仅返回标准JSON格式:{\"country\": [\"country\"], \"city\": [\"\
city\"]}\n - 无地理信息时返回:{\"country\": [], \"city\": []}\n - 禁止包含任何解释性文字或XML标签\n\
\ - 翻译成英文\n5. 验证逻辑:\n - 确保城市隶属于国家(如东京属于日本)\n - 拒绝推测性结论,必须有文本依据\n\
\n注意事项:\n- 国家城市名称必须使用新闻原文中的语言版本\n- 不存在的行政关系需视为无效信息(如\"巴黎属于德国\"\n- 多地点新闻只提取核心事件发生地\n\
- 返回的国家和城市名称必须是通俗意义的常见的英文名称,不能输出国家和城市中文名称\n\n<additional_rules>\n1. 国家城市名称标准化:\n\
\ - 优先使用主权国家名称(如\"中国\"而非\"中国大陆\"\n "
- id: dfd9b8ea-7e8c-4f38-be89-75b1624b41a7
role: user
text: '{{#1727233123593.msg_info#}}}'
selected: false
title: LLM
type: llm
variables: []
vision:
enabled: false
height: 89
id: '1745905312976'
position:
x: 317.04986301192474
y: 296.0789646906747
positionAbsolute:
x: 317.04986301192474
y: 296.0789646906747
selected: false
sourcePosition: right
targetPosition: left
type: custom
width: 243
viewport:
x: 166.56134424787098
y: 230.7781269141013
zoom: 0.25
app:
description: 新闻标签
icon: ✉️
icon_background: '#E0F2FE'
mode: workflow
name: 提取新闻发生的事件及经纬度信息-DeepseekV3
use_icon_as_answer_icon: false
dependencies:
- current_identifier: null
type: marketplace
value:
marketplace_plugin_unique_identifier: langgenius/volcengine_maas:0.0.10@ba0a7fc3151e89bf8645cd3abb3ccff971ee59e5b1127efca50b092dbb93bf80
kind: app
version: 0.1.5
workflow:
conversation_variables: []
environment_variables: []
features:
file_upload:
allowed_file_extensions:
- .JPG
- .JPEG
- .PNG
- .GIF
- .WEBP
- .SVG
allowed_file_types:
- image
allowed_file_upload_methods:
- local_file
- remote_url
enabled: false
fileUploadConfig:
audio_file_size_limit: 50
batch_count_limit: 5
file_size_limit: 15
image_file_size_limit: 10
video_file_size_limit: 100
workflow_file_upload_limit: 10
image:
enabled: false
number_limits: 3
transfer_methods:
- local_file
- remote_url
number_limits: 3
opening_statement: ''
retriever_resource:
enabled: false
sensitive_word_avoidance:
enabled: false
speech_to_text:
enabled: false
suggested_questions: []
suggested_questions_after_answer:
enabled: false
text_to_speech:
enabled: false
language: ''
voice: ''
graph:
edges:
- data:
isInLoop: false
sourceType: code
targetType: end
id: 1745896611748-source-1727235861736-target
source: '1745896611748'
sourceHandle: source
target: '1727235861736'
targetHandle: target
type: custom
zIndex: 0
- data:
isInIteration: false
isInLoop: false
sourceType: start
targetType: llm
id: 1727233123593-source-1745905312976-target
source: '1727233123593'
sourceHandle: source
target: '1745905312976'
targetHandle: target
type: custom
zIndex: 0
- data:
isInLoop: false
sourceType: llm
targetType: code
id: 1745905312976-source-1745896611748-target
source: '1745905312976'
sourceHandle: source
target: '1745896611748'
targetHandle: target
type: custom
zIndex: 0
nodes:
- data:
desc: "{\n \"msg_info\": \"DeryckWhibley坚称他'不是骗子',前经理否认了性胁迫指控\",\n \"sys.files\"\
: [],\n \"sys.user_id\": \"sl_backend_twitter\"\n}"
selected: false
title: 开始
type: start
variables:
- label: msg_info
max_length: 2000
options: []
required: true
type: paragraph
variable: msg_info
height: 197
id: '1727233123593'
position:
x: -36.369609408540725
y: 274.1424650967822
positionAbsolute:
x: -36.369609408540725
y: 274.1424650967822
selected: false
sourcePosition: right
targetPosition: left
type: custom
width: 243
- data:
desc: ''
outputs:
- value_selector:
- '1745896611748'
- result
variable: result
selected: false
title: 结束
type: end
height: 89
id: '1727235861736'
position:
x: 1052.254623008516
y: 310.3874923435948
positionAbsolute:
x: 1052.254623008516
y: 310.3874923435948
selected: false
sourcePosition: right
targetPosition: left
type: custom
width: 243
- data:
code: "def main(text: str) -> dict:\n json_part = text.split(\"</think>\"\
)[1].replace(\"```json\", \"\").replace(\"```\", \"\").replace('\\\\n',\
\ '').strip().replace('\\\\','')\n data = json.loads(json_part)\n \
\ r = []\n r.append(data)\n return {\n \"result\": r,\n \
\ }\n"
code_language: python3
desc: ''
outputs:
result:
children: null
type: array[object]
selected: false
title: 代码执行
type: code
variables:
- value_selector:
- '1745905312976'
- text
variable: text
height: 53
id: '1745896611748'
position:
x: 689.7058009416385
y: 318.0154642845672
positionAbsolute:
x: 689.7058009416385
y: 318.0154642845672
selected: false
sourcePosition: right
targetPosition: left
type: custom
width: 243
- data:
context:
enabled: true
variable_selector:
- '1727233123593'
- msg_info
desc: ''
model:
completion_params:
temperature: 0.7
mode: chat
name: Deepseek-R1
provider: langgenius/volcengine_maas/volcengine_maas
prompt_template:
- id: e49c99a8-1450-4a33-aac4-ab525eb3e379
role: system
text: "```xml\n<instruction>\n你是一个专业的犯罪新闻事件及经纬度提取助手,能够精准地从各类文本中提取新闻犯罪事件的关键信息。请按照以下步骤处理输入的文本信息:\n\
\n1. **提取犯罪事件发生时间**:从文本中识别事件发生的具体时间,包括日期和时间(格式必须为“2023-05-15”)。若文本中未明确提及时间,则标记为“”,输出字段为:event_time。\n\
\n2. **分类犯罪事件类型**:从给定的犯罪类型列表中选择最匹配的类别。列表包括:纵火、袭击、入室盗窃、扰乱公共秩序、毒品/酒精、酒驾/醉驾、欺诈、杀人、车辆盗窃、抢劫、性犯罪、盗窃、破坏公物、车内盗窃、武器犯罪、其他。若无法明确分类,则标记为“其他”,输出字段为:event_type。\n\
\n3. **确定犯罪点经纬度**:从文本中提取事件发生的具体地点(如“Bang Lamung, Chon Buri”)且不需要翻译,并转换为经纬度坐标。若地点不明确,则采用最匹配地点的经纬度(如城市中心坐标)。若无法确定,则标记为“”,输出字段为:location,longitude,latitude。\n\
\n4. **总结犯罪详情**:用50字以内的简洁语言概括事件的核心内容,包括犯罪手段、受害者或损失情况等,输出字段为:content。\n\
\n请按照以下步骤操作:\n- 仔细阅读输入的消息,确保理解所有关键信息。。\n- 用简洁的语言总结消息,确保通俗易懂且不超过50字。\n\
- 根据事件性质判断事件类型(如纵火、袭击、入室盗窃、扰乱公共秩序等)。\n- 将提取的信息整理为JSON格式,确保字段名与要求一致(event_time、event_type、location、longitude、latitude、content)。\n\
- 事件发生的时间格式,必须为:xxxx-xx-xx。\n- 输出时不要包含任何XML标签,仅提供纯JSON格式内容。\n输出格式为JSON,包含以下字段:\
\ \n- `event_time`:事件时间 \n- `event_type`:事件类型 \n- `location`:事件发生地点\n\
- `longitude`:经度 \n- `latitude`:纬度 \n- `content`:事件详情 \n</instruction>\n\
\n<examples>\n<example>\n输入: \n“2023年10月12日晚,一名男子在上海市浦东新区持刀抢劫便利店,抢走现金5000元后逃离现场。”\
\ \n\n输出: \n{\n \"event_time\": \"2023-10-12\",\n \"event_type\":\
\ \"抢劫\",\n \"location\": \"上海市浦东新区\",\n \"longitude\": \"121.5677\"\
,\n \"latitude\": \"31.2456\",\n \"content\": \"男子持刀抢劫便利店,抢走现金5000元。\"\
\n}\n</example>\n\n<example>\n输入: \n“5月20日凌晨,杭州市西湖区某酒吧发生群体斗殴事件,造成3人轻伤,警方已介入调查。”\
\ \n\n输出: \n{\n \"event_time\": \"2025-05-20\",\n \"event_type\":\
\ \"扰乱公共秩序\",\n \"location\": \"杭州市西湖区\",\n \"longitude\": \"\n \"\
latitude\": \"\",\n \"content\": \"杭州市西湖区某酒吧发生群体斗殴事件\"\n}\n</example>"
- id: dfd9b8ea-7e8c-4f38-be89-75b1624b41a7
role: user
text: '{{#1727233123593.msg_info#}}}'
selected: true
title: LLM
type: llm
variables: []
vision:
enabled: false
height: 89
id: '1745905312976'
position:
x: 321.58798976900476
y: 293.2033137258802
positionAbsolute:
x: 321.58798976900476
y: 293.2033137258802
selected: true
sourcePosition: right
targetPosition: left
type: custom
width: 243
viewport:
x: -270.50630049668894
y: 9.884318461029522
zoom: 0.8242832714190462
app:
description: 新闻标签
icon: ✉️
icon_background: '#E0F2FE'
mode: workflow
name: 提取紧急通知事件和总结-DeepseekV3
use_icon_as_answer_icon: false
dependencies:
- current_identifier: null
type: marketplace
value:
marketplace_plugin_unique_identifier: langgenius/volcengine_maas:0.0.10@ba0a7fc3151e89bf8645cd3abb3ccff971ee59e5b1127efca50b092dbb93bf80
kind: app
version: 0.1.5
workflow:
conversation_variables: []
environment_variables: []
features:
file_upload:
allowed_file_extensions:
- .JPG
- .JPEG
- .PNG
- .GIF
- .WEBP
- .SVG
allowed_file_types:
- image
allowed_file_upload_methods:
- local_file
- remote_url
enabled: false
fileUploadConfig:
audio_file_size_limit: 50
batch_count_limit: 5
file_size_limit: 15
image_file_size_limit: 10
video_file_size_limit: 100
workflow_file_upload_limit: 10
image:
enabled: false
number_limits: 3
transfer_methods:
- local_file
- remote_url
number_limits: 3
opening_statement: ''
retriever_resource:
enabled: false
sensitive_word_avoidance:
enabled: false
speech_to_text:
enabled: false
suggested_questions: []
suggested_questions_after_answer:
enabled: false
text_to_speech:
enabled: false
language: ''
voice: ''
graph:
edges:
- data:
isInIteration: false
sourceType: start
targetType: llm
id: 1727233123593-source-1732778846154-target
selected: false
source: '1727233123593'
sourceHandle: source
target: '1732778846154'
targetHandle: target
type: custom
zIndex: 0
- data:
isInIteration: false
isInLoop: false
sourceType: llm
targetType: code
id: 1732778846154-source-1746776906143-target
source: '1732778846154'
sourceHandle: source
target: '1746776906143'
targetHandle: target
type: custom
zIndex: 0
- data:
isInIteration: false
isInLoop: false
sourceType: code
targetType: end
id: 1746776906143-source-1727235861736-target
source: '1746776906143'
sourceHandle: source
target: '1727235861736'
targetHandle: target
type: custom
zIndex: 0
nodes:
- data:
desc: "{\n \"msg_info\": \"DeryckWhibley坚称他'不是骗子',前经理否认了性胁迫指控\",\n \"sys.files\"\
: [],\n \"sys.user_id\": \"sl_backend_twitter\"\n}"
selected: false
title: 开始
type: start
variables:
- label: msg_info
max_length: 2000
options: []
required: true
type: paragraph
variable: msg_info
height: 197
id: '1727233123593'
position:
x: -36.369609408540725
y: 274.1424650967822
positionAbsolute:
x: -36.369609408540725
y: 274.1424650967822
selected: false
sourcePosition: right
targetPosition: left
type: custom
width: 243
- data:
desc: ''
outputs:
- value_selector:
- '1746776906143'
- result
variable: result
selected: true
title: 结束
type: end
height: 89
id: '1727235861736'
position:
x: 1682.6905389121184
y: 312.1357764408275
positionAbsolute:
x: 1682.6905389121184
y: 312.1357764408275
selected: true
sourcePosition: right
targetPosition: left
type: custom
width: 243
- data:
context:
enabled: false
variable_selector: []
desc: ''
model:
completion_params:
top_p: 0.7
mode: chat
name: Deepseek-R1
provider: langgenius/volcengine_maas/volcengine_maas
prompt_template:
- id: ff06405d-dfb9-40ca-987a-c24fd3cfd6fb
role: system
text: "```xml\n<instruction>\n根据文旅部及外交部紧急通知消息,你需要完成以下任务:\n1. 提炼关键信息并输出指定格式的JSON。\n\
2. 提炼相关主体事件(无需考虑使馆提醒部分,但保留时间),输出字段为summary。\n3. 提取该事件的注意事项(尽量控制在50字内,用通俗的单句表达),输出字段为note。\n\
4. 提取事件类型,输出字段为event。\n\n请按照以下步骤操作:\n- 仔细阅读输入的通知消息,确保理解所有关键信息。\n- 从通知中提取与主体事件直接相关的内容,包括时间、地点、事件描述等,但忽略使馆提醒部分。\n\
- 用简洁的语言总结注意事项,确保通俗易懂且不超过50字。\n- 根据事件性质判断事件类型(如自然灾害、政治事件、公共卫生事件等),仅提取一件事件。\n\
- 将提取的信息整理为JSON格式,确保字段名与要求一致(summary、note、event)。\n- 输出时不要包含任何XML标签,仅提供纯JSON格式内容。\n\
\n请确保输出内容准确、简洁,且完全基于输入信息。不要添加任何未提及的细节或推测。\n</instruction>\n\n<examples>\n\
<example>\n<input>\n文旅部及外交部紧急通知:近日,某国发生大规模示威活动,部分地区出现暴力冲突。使馆提醒中国公民避免前往示威区域。\n\
</input>\n<output>\n{\n \"summary\": \"某国近日发生大规模示威活动,部分地区出现暴力冲突\",\n\
\ \"note\": \"避免前往示威区域以防卷入冲突\",\n \"event\": \"政治事件\"\n}\n</output>\n\
</example>\n</examples>"
- id: 9f485eef-bc41-4f14-b8dd-9d507b6a8d3d
role: user
text: '{{#1727233123593.msg_info#}}'
selected: false
title: LLM
type: llm
variables: []
vision:
enabled: false
height: 89
id: '1732778846154'
position:
x: 625.935426523185
y: 303.6520740128239
positionAbsolute:
x: 625.935426523185
y: 303.6520740128239
selected: false
sourcePosition: right
targetPosition: left
type: custom
width: 243
- data:
code: "def main(text: str) -> dict:\n json_part = text.split(\"</think>\"\
)[1].replace(\"```json\", \"\").replace(\"```\", \"\").replace('\\\\n',\
\ '').strip().replace('\\\\','')\n data = json.loads(json_part)\n \
\ r = []\n r.append(data)\n return {\n \"result\": r,\n \
\ }"
code_language: python3
desc: ''
outputs:
result:
children: null
type: array[object]
selected: false
title: 代码执行
type: code
variables:
- value_selector:
- '1732778846154'
- text
variable: text
height: 53
id: '1746776906143'
position:
x: 1120.1084646683125
y: 312.1357764408275
positionAbsolute:
x: 1120.1084646683125
y: 312.1357764408275
selected: false
sourcePosition: right
targetPosition: left
type: custom
width: 243
viewport:
x: 267.53983034654414
y: 352.5777526112109
zoom: 0.30759718263110325
app:
description: 新闻标签
icon: ✉️
icon_background: '#E0F2FE'
mode: workflow
name: 提取紧急通知信息-DeepseekV3
use_icon_as_answer_icon: false
dependencies:
- current_identifier: null
type: marketplace
value:
marketplace_plugin_unique_identifier: langgenius/volcengine_maas:0.0.10@ba0a7fc3151e89bf8645cd3abb3ccff971ee59e5b1127efca50b092dbb93bf80
kind: app
version: 0.1.5
workflow:
conversation_variables: []
environment_variables: []
features:
file_upload:
allowed_file_extensions:
- .JPG
- .JPEG
- .PNG
- .GIF
- .WEBP
- .SVG
allowed_file_types:
- image
allowed_file_upload_methods:
- local_file
- remote_url
enabled: false
fileUploadConfig:
audio_file_size_limit: 50
batch_count_limit: 5
file_size_limit: 15
image_file_size_limit: 10
video_file_size_limit: 100
workflow_file_upload_limit: 10
image:
enabled: false
number_limits: 3
transfer_methods:
- local_file
- remote_url
number_limits: 3
opening_statement: ''
retriever_resource:
enabled: false
sensitive_word_avoidance:
enabled: false
speech_to_text:
enabled: false
suggested_questions: []
suggested_questions_after_answer:
enabled: false
text_to_speech:
enabled: false
language: ''
voice: ''
graph:
edges:
- data:
isInIteration: false
sourceType: start
targetType: llm
id: 1727233123593-source-1732778846154-target
source: '1727233123593'
sourceHandle: source
target: '1732778846154'
targetHandle: target
type: custom
zIndex: 0
- data:
isInLoop: false
sourceType: llm
targetType: code
id: 1732778846154-source-1745387489949-target
source: '1732778846154'
sourceHandle: source
target: '1745387489949'
targetHandle: target
type: custom
zIndex: 0
- data:
isInLoop: false
sourceType: code
targetType: end
id: 1745387489949-source-1727235861736-target
source: '1745387489949'
sourceHandle: source
target: '1727235861736'
targetHandle: target
type: custom
zIndex: 0
nodes:
- data:
desc: "{\n \"msg_info\": \"DeryckWhibley坚称他'不是骗子',前经理否认了性胁迫指控\",\n \"sys.files\"\
: [],\n \"sys.user_id\": \"sl_backend_twitter\"\n}"
selected: false
title: 开始
type: start
variables:
- label: msg_info
max_length: 2000
options: []
required: true
type: paragraph
variable: msg_info
height: 197
id: '1727233123593'
position:
x: -36.369609408540725
y: 274.1424650967822
positionAbsolute:
x: -36.369609408540725
y: 274.1424650967822
selected: false
sourcePosition: right
targetPosition: left
type: custom
width: 243
- data:
desc: ''
outputs:
- value_selector:
- '1745387489949'
- result
variable: result
selected: false
title: 结束
type: end
height: 89
id: '1727235861736'
position:
x: 1179.4863206530927
y: 314.7747922623733
positionAbsolute:
x: 1179.4863206530927
y: 314.7747922623733
selected: false
sourcePosition: right
targetPosition: left
type: custom
width: 243
- data:
context:
enabled: false
variable_selector: []
desc: ''
model:
completion_params:
top_p: 0.7
mode: chat
name: deepseek-v3-250324
provider: langgenius/volcengine_maas/volcengine_maas
prompt_template:
- id: ff06405d-dfb9-40ca-987a-c24fd3cfd6fb
role: system
text: "```xml\n<instruction>\n根据给定的多条JSON格式的文旅部及外交部紧急通知消息,每条消息包含id、title和content字段。你需要完成以下任务:\n\
\n1. 紧急程度判断:根据title和content的内容,按照以下标准判断紧急程度:\n - 极高:存在致命威胁且需强制撤离\n \
\ - 高:存在重大安全威胁需规避行程\n - 中:局部或特定风险需加强防范\n - 低:常规安全提醒\n - 极低:日常注意事项\n\
\n2. 区域判断:判断消息描述的是境内还是境外(相对中国而言):\n - 境内:描述中国国内的情况\n - 境外:描述中国以外国家或地区的情况\n\
\n3. 地理位置提取:从内容中提取描述的国家和城市:\n - 国家:必须提取,如果是境内则为\"中国\"\n - 城市:尽可能提取,无法提取则输出\"\
\",多个城市以英文,分割\n\n4. 输出格式要求:\n - 保持原始id不变\n - 新增字段:\n * level:表示紧急程度\n\
\ * overseas:表示境内/境外\n * country:表示国家\n * city:表示城市\n -\
\ 输出为JSON数组格式\n - 不要包含任何XML标签\n\n处理步骤:\n1. 仔细阅读每条消息的title和content\n\
2. 根据内容严重性判断紧急程度\n3. 判断是否涉及中国境内\n4. 提取明确提到的国家和城市\n5. 按照要求格式组织输出\n\n注意事项:\n\
- 城市提取要准确,不确定时留空\n- 紧急程度判断要严格遵循标准\n- 确保输出是标准JSON格式\n</instruction>\n\n\
<examples>\n示例1:\n输入:\n[\n {\n \"id\": 1,\n \"title\": \"关于暂勿前往XX国的紧急提醒\"\
,\n \"content\": \"近期XX国发生武装冲突,已造成多人伤亡,建议中国公民暂勿前往。特别提醒在X市的中国公民尽快撤离。\"\
\n }\n]\n输出:\n[\n {\n \"id\": 1,\n \"level\": \"极高\",\n \"\
area\": \"境外\",\n \"country\": \"XX国\",\n \"city\": \"X市\"\n }\n\
]\n</instruction>\n```"
- id: 9f485eef-bc41-4f14-b8dd-9d507b6a8d3d
role: user
text: '{{#1727233123593.msg_info#}}'
selected: true
title: LLM
type: llm
variables: []
vision:
enabled: false
height: 89
id: '1732778846154'
position:
x: 389.99627588072235
y: 320.20911424395644
positionAbsolute:
x: 389.99627588072235
y: 320.20911424395644
selected: true
sourcePosition: right
targetPosition: left
type: custom
width: 243
- data:
code: "\ndef main(text: str) -> dict:\n text = text.replace(\"输出:\",\"\"\
)\n text = text.replace(\"```json\", \"\").replace(\"```\", \"\").strip()\n\
\ data = json.loads(text)\n return {\n \"result\": data,\n\
\ }\n\n"
code_language: python3
desc: ''
outputs:
result:
children: null
type: array[object]
selected: false
title: 代码执行
type: code
variables:
- value_selector:
- '1732778846154'
- text
variable: text
height: 53
id: '1745387489949'
position:
x: 754.4482729679055
y: 320.20911424395644
positionAbsolute:
x: 754.4482729679055
y: 320.20911424395644
selected: false
sourcePosition: right
targetPosition: left
type: custom
width: 243
viewport:
x: 105.15857764434577
y: 299.2134077214364
zoom: 0.5236470614103135
This diff is collapsed.
This diff is collapsed.
app:
description: 规范自然灾害事件
icon: ✉️
icon_background: '#E0F2FE'
mode: workflow
name: 消息标签-大模型v0.2.0
use_icon_as_answer_icon: false
dependencies:
- current_identifier: null
type: marketplace
value:
marketplace_plugin_unique_identifier: langgenius/deepseek:0.0.5@fd6efd37c2a931911de8ab9ca3ba2da303bef146d45ee87ad896b04b36d09403
kind: app
version: 0.1.5
workflow:
conversation_variables: []
environment_variables: []
features:
file_upload:
allowed_file_extensions:
- .JPG
- .JPEG
- .PNG
- .GIF
- .WEBP
- .SVG
allowed_file_types:
- image
allowed_file_upload_methods:
- local_file
- remote_url
enabled: false
fileUploadConfig:
audio_file_size_limit: 50
batch_count_limit: 5
file_size_limit: 15
image_file_size_limit: 10
video_file_size_limit: 100
workflow_file_upload_limit: 10
image:
enabled: false
number_limits: 3
transfer_methods:
- local_file
- remote_url
number_limits: 3
opening_statement: ''
retriever_resource:
enabled: false
sensitive_word_avoidance:
enabled: false
speech_to_text:
enabled: false
suggested_questions: []
suggested_questions_after_answer:
enabled: false
text_to_speech:
enabled: false
language: ''
voice: ''
graph:
edges:
- data:
isInIteration: false
sourceType: start
targetType: llm
id: 1727233123593-source-1743569567393-target
source: '1727233123593'
sourceHandle: source
target: '1743569567393'
targetHandle: target
type: custom
zIndex: 0
- data:
isInIteration: false
sourceType: llm
targetType: end
id: 1743569567393-source-1727235861736-target
source: '1743569567393'
sourceHandle: source
target: '1727235861736'
targetHandle: target
type: custom
zIndex: 0
nodes:
- data:
desc: "{\n \"msg_info\": \"DeryckWhibley坚称他'不是骗子',前经理否认了性胁迫指控\",\n \"sys.files\"\
: [],\n \"sys.user_id\": \"sl_backend_twitter\"\n}"
selected: false
title: 开始
type: start
variables:
- label: msg_info
max_length: 2000
options: []
required: true
type: paragraph
variable: msg_info
height: 197
id: '1727233123593'
position:
x: -87.81948835004954
y: 205.81936206208348
positionAbsolute:
x: -87.81948835004954
y: 205.81936206208348
selected: false
sourcePosition: right
targetPosition: left
type: custom
width: 243
- data:
desc: ''
outputs:
- value_selector:
- '1727242813814'
- result
variable: result
selected: false
title: 结束
type: end
height: 89
id: '1727235861736'
position:
x: 950.3329417497814
y: 233.49376834959315
positionAbsolute:
x: 950.3329417497814
y: 233.49376834959315
selected: false
sourcePosition: right
targetPosition: left
type: custom
width: 243
- data:
context:
enabled: false
variable_selector: []
desc: ''
model:
completion_params:
temperature: 0.7
mode: chat
name: deepseek-chat
provider: langgenius/deepseek/deepseek
prompt_template:
- id: a0d77e73-c56f-4a1a-9046-ea4c3ec06709
role: system
text: '```xml
<instruction>
对多条新闻内容进行打标签,标签分别为枪击、抢劫、爆炸、恐怖袭击、纵火、大规模伤亡事件、人质事件、毒气泄漏、自然灾害、生物危害、交通意外、建筑物倒塌、食品安全、社会冲突、群体性事件等。请按照以下步骤完成任务:
1. 仔细阅读每条新闻内容,理解其核心事件。
2. 根据新闻描述的事件类型,从给定的标签列表中选择最匹配的一个或多个标签。
3. 确保标签与新闻内容高度相关,避免主观臆断。
4. 如果新闻内容涉及多个事件类型,可以分配多个标签,但需确保每个标签都准确反映新闻内容。
5. 输出结果时,仅列出标签名称,不要包含任何XML标签或其他无关内容。
<examples>
<example>
<输入>
"昨日凌晨,某市一家银行发生持枪抢劫案,歹徒与警方交火后逃离现场,造成两名保安受伤。"
<输出>
枪击, 抢劫
</example>
<example>
<输入>
"今日上午,某化工厂发生毒气泄漏事故,附近居民已被紧急疏散,暂无人员伤亡报告。"
<输出>
毒气泄漏
</example>
<example>
<输入>
"近日,某地因征地问题引发大规模抗议活动,数百名民众与警方发生冲突,导致多人受伤。"
<输出>
社会冲突, 群体性事件
</example>
</examples>
注意事项:
- 标签必须严格从给定的列表中选择,不得自行添加或修改标签名称。
- 如果新闻内容与任何标签都不匹配,则输出""。
- 输出时标签之间用逗号分隔,不要使用编号或项目符号。
- 确保标签的准确性和一致性,避免歧义或模糊分类。
请严格按照上述要求完成任务,确保输出结果简洁、准确且符合规范。
</instruction>
```'
- id: 79e4db09-9e63-4b9b-a599-5269c92ca0f9
role: user
text: '{{#1727233123593.msg_info#}}'
selected: true
title: LLM
type: llm
variables: []
vision:
enabled: false
height: 89
id: '1743569567393'
position:
x: 407.7152535151089
y: 224.80166055583152
positionAbsolute:
x: 407.7152535151089
y: 224.80166055583152
selected: true
sourcePosition: right
targetPosition: left
type: custom
width: 243
viewport:
x: 238.23781297303276
y: 137.95720494469128
zoom: 1.1062938456539644
app:
description: 新闻标签
icon: ✉️
icon_background: '#E0F2FE'
mode: workflow
name: 电源信息结构化-DeepseekV3
use_icon_as_answer_icon: false
dependencies:
- current_identifier: null
type: marketplace
value:
marketplace_plugin_unique_identifier: langgenius/volcengine_maas:0.0.10@ba0a7fc3151e89bf8645cd3abb3ccff971ee59e5b1127efca50b092dbb93bf80
kind: app
version: 0.1.5
workflow:
conversation_variables: []
environment_variables: []
features:
file_upload:
allowed_file_extensions:
- .JPG
- .JPEG
- .PNG
- .GIF
- .WEBP
- .SVG
allowed_file_types:
- image
allowed_file_upload_methods:
- local_file
- remote_url
enabled: false
fileUploadConfig:
audio_file_size_limit: 50
batch_count_limit: 5
file_size_limit: 15
image_file_size_limit: 10
video_file_size_limit: 100
workflow_file_upload_limit: 10
image:
enabled: false
number_limits: 3
transfer_methods:
- local_file
- remote_url
number_limits: 3
opening_statement: ''
retriever_resource:
enabled: false
sensitive_word_avoidance:
enabled: false
speech_to_text:
enabled: false
suggested_questions: []
suggested_questions_after_answer:
enabled: false
text_to_speech:
enabled: false
language: ''
voice: ''
graph:
edges:
- data:
isInIteration: false
sourceType: start
targetType: llm
id: 1727233123593-source-1732778846154-target
selected: false
source: '1727233123593'
sourceHandle: source
target: '1732778846154'
targetHandle: target
type: custom
zIndex: 0
- data:
isInIteration: false
isInLoop: false
sourceType: llm
targetType: code
id: 1732778846154-source-1748241838331-target
source: '1732778846154'
sourceHandle: source
target: '1748241838331'
targetHandle: target
type: custom
zIndex: 0
- data:
isInIteration: false
isInLoop: false
sourceType: code
targetType: end
id: 1748241838331-source-1727235861736-target
source: '1748241838331'
sourceHandle: source
target: '1727235861736'
targetHandle: target
type: custom
zIndex: 0
nodes:
- data:
desc: "{\n \"msg_info\": \"DeryckWhibley坚称他'不是骗子',前经理否认了性胁迫指控\",\n \"sys.files\"\
: [],\n \"sys.user_id\": \"sl_backend_twitter\"\n}"
selected: false
title: 开始
type: start
variables:
- label: msg_info
max_length: 2000
options: []
required: true
type: paragraph
variable: msg_info
height: 197
id: '1727233123593'
position:
x: -36.369609408540725
y: 274.1424650967822
positionAbsolute:
x: -36.369609408540725
y: 274.1424650967822
selected: false
sourcePosition: right
targetPosition: left
type: custom
width: 243
- data:
desc: ''
outputs:
- value_selector:
- '1748241838331'
- result
variable: result
selected: false
title: 结束
type: end
height: 89
id: '1727235861736'
position:
x: 1192.3926665576485
y: 328.2744664355823
positionAbsolute:
x: 1192.3926665576485
y: 328.2744664355823
selected: false
sourcePosition: right
targetPosition: left
type: custom
width: 243
- data:
context:
enabled: false
variable_selector: []
desc: ''
model:
completion_params:
top_p: 0.7
mode: chat
name: deepseek-v3-250324
provider: langgenius/volcengine_maas/volcengine_maas
prompt_template:
- id: ff06405d-dfb9-40ca-987a-c24fd3cfd6fb
role: system
text: "```xml\n<instruction>\n你是一个专业的数据筛选总结归纳者,请根据提供的文本内容,按照以下结构化格式进行总结归纳,并输出为JSON格式。具体要求如下:\n\
\n1. 严格按照以下7个分类进行归纳:\n a. 插头样式与电压(包含插头类型、电压)\n b. 主要通信运营商(列出所有运营商及联系方式)\n\
\ c. 网络类型支持(包含4G/5G覆盖范围及频段信息)\n d. 移动通信制式(明确4G/5G)\n e. 国际电话拨打规则(包含双向拨打格式)\n\
\ f. 中国国际漫游使用(运营商合作信息)\n g. 总结建议(给出实用建议)\n\n2. 处理规则:\n - 对原文明确的信息直接提取\n\
\ - 对未提及的信息标注\"未提及\"\n - 主要通信运营商联系电话号码未提及无需处理,直接不显示\n\n3. 输出格式要求:\n\
\ - 使用标准JSON数组格式\n - 每个分类作为独立对象\n - 对象包含\"key\"\"value\"字段\n \
\ - value字段使用数组格式存储多条信息\n\n4. 特别注意事项:\n - 不得修改原文数据\n - 保持数值和单位的完整性\n\
\ - 中英文标点使用要规范\n - 不要包含任何XML标签\n - 缺失信息需明确标注\n\n5. 操作步骤:\n - \
\ \"网络类型支持\"中的4G覆盖范围、5G覆盖范围分别总结为独立条目\n - \"移动通信制式\"中的4G、5G分别总结为独立条目\n\
\ - \"总结建议\"中的电力适配、通信选择、网络使用、资费提示、通话建议、漫游提醒分别总结为独立条目\n</instruction>\n\
\n<examples>\n示例1:\n输入:\"该国使用G型插头,电压标准为230V。主要运营商有A公司(12345)和B公司(67890)。4G覆盖主要城市。\"\
\n输出:\n[\n {\n \"key\": \"插头样式与电压\",\n \"value\": [\"插头类型:G型三脚插头\"\
, \"电压:230V\"]\n },\n {\n \"key\": \"主要通信运营商\",\n \"value\": [\"\
A公司:12345\", \"B公司:67890\"]\n },\n {\n \"key\": \"网络类型支持\",\n \
\ \"value\": [\"4G覆盖范围:主要城市\", \"5G覆盖范围:未提及\", \"频段:未提及\"]\n }\n]"
- id: 9f485eef-bc41-4f14-b8dd-9d507b6a8d3d
role: user
text: '{{#1727233123593.msg_info#}}'
selected: true
title: LLM
type: llm
variables: []
vision:
enabled: false
height: 89
id: '1732778846154'
position:
x: 349.0335068458452
y: 328.2744664355823
positionAbsolute:
x: 349.0335068458452
y: 328.2744664355823
selected: true
sourcePosition: right
targetPosition: left
type: custom
width: 243
- data:
code: "def main(text: str) -> dict:\n json_part = text.replace(\"```json\"\
, \"\").replace(\"```\", \"\").replace('\\\\n', '').strip().replace('\\\\\
','')\n data = json.loads(json_part)\n return {\n \"result\"\
: data,\n }"
code_language: python3
desc: ''
outputs:
result:
children: null
type: array[object]
selected: false
title: 代码执行
type: code
variables:
- value_selector:
- '1732778846154'
- text
variable: text
height: 53
id: '1748241838331'
position:
x: 734.3330920867545
y: 328.2744664355823
positionAbsolute:
x: 734.3330920867545
y: 328.2744664355823
selected: false
sourcePosition: right
targetPosition: left
type: custom
width: 243
viewport:
x: 225.5621639208738
y: 319.1090901244189
zoom: 0.7085935817233538
app:
description: 新闻标签
icon: ✉️
icon_background: '#E0F2FE'
mode: workflow
name: 紧急联络结构化-DeepseekV3
use_icon_as_answer_icon: false
dependencies:
- current_identifier: null
type: marketplace
value:
marketplace_plugin_unique_identifier: langgenius/volcengine_maas:0.0.10@ba0a7fc3151e89bf8645cd3abb3ccff971ee59e5b1127efca50b092dbb93bf80
kind: app
version: 0.1.5
workflow:
conversation_variables: []
environment_variables: []
features:
file_upload:
allowed_file_extensions:
- .JPG
- .JPEG
- .PNG
- .GIF
- .WEBP
- .SVG
allowed_file_types:
- image
allowed_file_upload_methods:
- local_file
- remote_url
enabled: false
fileUploadConfig:
audio_file_size_limit: 50
batch_count_limit: 5
file_size_limit: 15
image_file_size_limit: 10
video_file_size_limit: 100
workflow_file_upload_limit: 10
image:
enabled: false
number_limits: 3
transfer_methods:
- local_file
- remote_url
number_limits: 3
opening_statement: ''
retriever_resource:
enabled: false
sensitive_word_avoidance:
enabled: false
speech_to_text:
enabled: false
suggested_questions: []
suggested_questions_after_answer:
enabled: false
text_to_speech:
enabled: false
language: ''
voice: ''
graph:
edges:
- data:
isInIteration: false
sourceType: start
targetType: llm
id: 1727233123593-source-1732778846154-target
selected: false
source: '1727233123593'
sourceHandle: source
target: '1732778846154'
targetHandle: target
type: custom
zIndex: 0
- data:
isInIteration: false
isInLoop: false
sourceType: llm
targetType: code
id: 1732778846154-source-1748241838331-target
source: '1732778846154'
sourceHandle: source
target: '1748241838331'
targetHandle: target
type: custom
zIndex: 0
- data:
isInIteration: false
isInLoop: false
sourceType: code
targetType: end
id: 1748241838331-source-1727235861736-target
source: '1748241838331'
sourceHandle: source
target: '1727235861736'
targetHandle: target
type: custom
zIndex: 0
nodes:
- data:
desc: "{\n \"msg_info\": \"DeryckWhibley坚称他'不是骗子',前经理否认了性胁迫指控\",\n \"sys.files\"\
: [],\n \"sys.user_id\": \"sl_backend_twitter\"\n}"
selected: false
title: 开始
type: start
variables:
- label: msg_info
max_length: 2000
options: []
required: true
type: paragraph
variable: msg_info
height: 197
id: '1727233123593'
position:
x: -36.369609408540725
y: 274.1424650967822
positionAbsolute:
x: -36.369609408540725
y: 274.1424650967822
selected: false
sourcePosition: right
targetPosition: left
type: custom
width: 243
- data:
desc: ''
outputs:
- value_selector:
- '1748241838331'
- result
variable: result
selected: false
title: 结束
type: end
height: 89
id: '1727235861736'
position:
x: 1192.3926665576485
y: 328.2744664355823
positionAbsolute:
x: 1192.3926665576485
y: 328.2744664355823
selected: false
sourcePosition: right
targetPosition: left
type: custom
width: 243
- data:
context:
enabled: false
variable_selector: []
desc: ''
model:
completion_params:
top_p: 0.7
mode: chat
name: deepseek-v3-250324
provider: langgenius/volcengine_maas/volcengine_maas
prompt_template:
- id: ff06405d-dfb9-40ca-987a-c24fd3cfd6fb
role: system
text: "```xml\n<instruction>\n你是一个专业的数据筛选和总结归纳者,你的任务是从提供的文本内容中提取关键电话联系方式,并按照指定的JSON格式进行整理输出。请严格遵循以下步骤:\n\
\n1. **仔细阅读输入文本**:确保理解所有提到的电话联系方式及其对应的类别。\n2. **分类整理**:将电话联系方式分为以下6类:\n\
\ - 紧急联络\n - 中国驻该国机构\n - 其他非常重要的出境须知电话\n3. **提取并验证**:确保每个电话号码的准确性和完整性,避免遗漏或错误。\n\
4. 操作步骤:\n(1) 首先提取原文中与三大分类相关的所有信息\n(2) 对每个大类下的细分条目进行语义压缩,确保核心信息完整,且细分只总结一条\n\
(3) \"紧急联络\"中的火警、匪警、急救、查号台分别总结为独立条目\n(4) \"中国驻该国机构\"中的使馆值班手机、领事保护应急电话、经济商务处、南方航空巴库办事处分别总结为独立条目\n\
(5) \"其他非常重要的出境须知电话\"中的紧急救援服务、移民局热线、出租车服务、电力供应紧急服、货币汇率查询、租赁汽车服务、ASAN政务服务中心热线、纳希切万外事局分别总结为独立条目\n\
5. **JSON格式输出**:按照以下格式输出结果,确保每个类别和对应的电话号码清晰列出:\n ```json\n [\n \
\ {\n \"key\": \"类别名称\",\n \"value\": [\"电话描述:电话号码\"]\n \
\ }\n ]\n ```\n6. **注意事项**:\n - 如果某个类别没有直接电话,请标注“无直接电话”。\n \
\ - 确保输出为纯JSON格式,不包含任何XML标签或其他无关内容。\n - 如果同一类别有多个电话,请全部列出。\n\n请确保输出的JSON格式正确无误,且内容完整。\n\
</instruction>"
- id: 9f485eef-bc41-4f14-b8dd-9d507b6a8d3d
role: user
text: '{{#1727233123593.msg_info#}}'
selected: true
title: LLM
type: llm
variables: []
vision:
enabled: false
height: 89
id: '1732778846154'
position:
x: 349.0335068458452
y: 328.2744664355823
positionAbsolute:
x: 349.0335068458452
y: 328.2744664355823
selected: true
sourcePosition: right
targetPosition: left
type: custom
width: 243
- data:
code: "def main(text: str) -> dict:\n json_part = text.replace(\"```json\"\
, \"\").replace(\"```\", \"\").replace('\\\\n', '').strip().replace('\\\\\
','')\n data = json.loads(json_part)\n return {\n \"result\"\
: data,\n }"
code_language: python3
desc: ''
outputs:
result:
children: null
type: array[object]
selected: false
title: 代码执行
type: code
variables:
- value_selector:
- '1732778846154'
- text
variable: text
height: 53
id: '1748241838331'
position:
x: 734.3330920867545
y: 328.2744664355823
positionAbsolute:
x: 734.3330920867545
y: 328.2744664355823
selected: false
sourcePosition: right
targetPosition: left
type: custom
width: 243
viewport:
x: 225.5621639208738
y: 319.1090901244189
zoom: 0.7085935817233538
app:
description: 新闻标签
icon: ✉️
icon_background: '#E0F2FE'
mode: workflow
name: 紧急通知周报-DeepseekR1
use_icon_as_answer_icon: false
dependencies:
- current_identifier: null
type: marketplace
value:
marketplace_plugin_unique_identifier: langgenius/volcengine_maas:0.0.10@ba0a7fc3151e89bf8645cd3abb3ccff971ee59e5b1127efca50b092dbb93bf80
kind: app
version: 0.1.5
workflow:
conversation_variables: []
environment_variables: []
features:
file_upload:
allowed_file_extensions:
- .JPG
- .JPEG
- .PNG
- .GIF
- .WEBP
- .SVG
allowed_file_types:
- image
allowed_file_upload_methods:
- local_file
- remote_url
enabled: false
fileUploadConfig:
audio_file_size_limit: 50
batch_count_limit: 5
file_size_limit: 15
image_file_size_limit: 10
video_file_size_limit: 100
workflow_file_upload_limit: 10
image:
enabled: false
number_limits: 3
transfer_methods:
- local_file
- remote_url
number_limits: 3
opening_statement: ''
retriever_resource:
enabled: false
sensitive_word_avoidance:
enabled: false
speech_to_text:
enabled: false
suggested_questions: []
suggested_questions_after_answer:
enabled: false
text_to_speech:
enabled: false
language: ''
voice: ''
graph:
edges:
- data:
isInIteration: false
sourceType: start
targetType: llm
id: 1727233123593-source-1732778846154-target
selected: false
source: '1727233123593'
sourceHandle: source
target: '1732778846154'
targetHandle: target
type: custom
zIndex: 0
- data:
isInIteration: false
isInLoop: false
sourceType: llm
targetType: code
id: 1732778846154-source-1746776906143-target
source: '1732778846154'
sourceHandle: source
target: '1746776906143'
targetHandle: target
type: custom
zIndex: 0
- data:
isInIteration: false
isInLoop: false
sourceType: code
targetType: end
id: 1746776906143-source-1727235861736-target
source: '1746776906143'
sourceHandle: source
target: '1727235861736'
targetHandle: target
type: custom
zIndex: 0
nodes:
- data:
desc: "{\n \"msg_info\": \"DeryckWhibley坚称他'不是骗子',前经理否认了性胁迫指控\",\n \"sys.files\"\
: [],\n \"sys.user_id\": \"sl_backend_twitter\"\n}"
selected: false
title: 开始
type: start
variables:
- label: msg_info
max_length: 2000
options: []
required: true
type: paragraph
variable: msg_info
height: 197
id: '1727233123593'
position:
x: -36.369609408540725
y: 274.1424650967822
positionAbsolute:
x: -36.369609408540725
y: 274.1424650967822
selected: false
sourcePosition: right
targetPosition: left
type: custom
width: 243
- data:
desc: ''
outputs:
- value_selector:
- '1746776906143'
- result
variable: result
selected: false
title: 结束
type: end
height: 89
id: '1727235861736'
position:
x: 1682.6905389121184
y: 312.1357764408275
positionAbsolute:
x: 1682.6905389121184
y: 312.1357764408275
selected: false
sourcePosition: right
targetPosition: left
type: custom
width: 243
- data:
context:
enabled: false
variable_selector: []
desc: ''
model:
completion_params:
top_p: 0.7
mode: chat
name: Deepseek-R1
provider: langgenius/volcengine_maas/volcengine_maas
prompt_template:
- id: ff06405d-dfb9-40ca-987a-c24fd3cfd6fb
role: system
text: '```xml
<instruction>
根据多条文旅部或外交部发布的紧急通知信息,生成一段简短的概括性安全提示。提示需涵盖保险责任相关内容,并以“Security Link在此提醒您”开头。具体要求如下:
1. 内容需简洁明了,突出保险责任和风险规避建议。
2. 语言风格需正式且具有警示性。
3. 不得包含任何XML标签。
4. 根据输入信息中的关键风险点(如示威游行、自然灾害等)生成对应的保险责任提示。
5. 结尾需提供具体的行动建议(如“避开相关区域”或“联系保险公司”等)。
生成步骤:
1. 提取输入信息中的核心风险事件(如战争、罢工、自然灾害等)。
2. 明确该风险事件在保险条款中的常见责任归属(如“除外责任”或“有限保障”)。
3. 组合成标准句式:“Security Link在此提醒您,[风险事件]常被列为[保险责任说明],因此[行动建议]。”
4. 检查语句通顺性和专业性,确保无歧义。
</instruction>
<examples>
<example>
<input>
外交部提醒:某国首都爆发大规模反政府示威,局部地区出现武装冲突
</input>
<output>
Security Link在此提醒您,示威游行及武装冲突事件通常被列为旅行保险的除外责任,建议您暂勿前往相关区域,并确认保单中的战争条款保障范围。
</output>
</example>
<example>
<input>
文旅部通知:B国沿海地区遭遇百年一遇台风,多趟航班取消
</input>
<output>
Security Link在此提醒您,自然灾害导致的行程中断可能需特定保险条款才予理赔,请立即联系保险公司报备,并保留所有交通票据作为理赔凭证。
</output>
</example>
<format_requirements>
1. 严格采用“Security Link在此提醒您”作为开头
2. 必须包含“保险”相关表述
3. 风险描述需与输入信息完全匹配
4. 行动建议需具体可行
5. 总字数控制在50-80字之间
'
- id: 9f485eef-bc41-4f14-b8dd-9d507b6a8d3d
role: user
text: '{{#1727233123593.msg_info#}}'
selected: false
title: LLM
type: llm
variables: []
vision:
enabled: false
height: 89
id: '1732778846154'
position:
x: 634.1052393737077
y: 305.69452722545464
positionAbsolute:
x: 634.1052393737077
y: 305.69452722545464
selected: false
sourcePosition: right
targetPosition: left
type: custom
width: 243
- data:
code: "def main(text: str) -> dict:\n data = text.split(\"</think>\")[1].replace('\\\
\\n', '').strip().replace('\\\\','')\n r = []\n summary={\n \
\ \"summary\":data\n }\n r.append(summary)\n return {\n \
\ \"result\": r,\n }"
code_language: python3
desc: ''
outputs:
result:
children: null
type: array[object]
selected: true
title: 代码执行
type: code
variables:
- value_selector:
- '1732778846154'
- text
variable: text
height: 53
id: '1746776906143'
position:
x: 1120.1084646683125
y: 312.1357764408275
positionAbsolute:
x: 1120.1084646683125
y: 312.1357764408275
selected: true
sourcePosition: right
targetPosition: left
type: custom
width: 243
viewport:
x: 50.56216392087401
y: 303.7901778011048
zoom: 0.7085935817233538
app:
description: 新闻标签
icon: ✉️
icon_background: '#E0F2FE'
mode: workflow
name: 风俗结构化-DeepseekV3
use_icon_as_answer_icon: false
dependencies:
- current_identifier: null
type: marketplace
value:
marketplace_plugin_unique_identifier: langgenius/volcengine_maas:0.0.10@ba0a7fc3151e89bf8645cd3abb3ccff971ee59e5b1127efca50b092dbb93bf80
kind: app
version: 0.1.5
workflow:
conversation_variables: []
environment_variables: []
features:
file_upload:
allowed_file_extensions:
- .JPG
- .JPEG
- .PNG
- .GIF
- .WEBP
- .SVG
allowed_file_types:
- image
allowed_file_upload_methods:
- local_file
- remote_url
enabled: false
fileUploadConfig:
audio_file_size_limit: 50
batch_count_limit: 5
file_size_limit: 15
image_file_size_limit: 10
video_file_size_limit: 100
workflow_file_upload_limit: 10
image:
enabled: false
number_limits: 3
transfer_methods:
- local_file
- remote_url
number_limits: 3
opening_statement: ''
retriever_resource:
enabled: false
sensitive_word_avoidance:
enabled: false
speech_to_text:
enabled: false
suggested_questions: []
suggested_questions_after_answer:
enabled: false
text_to_speech:
enabled: false
language: ''
voice: ''
graph:
edges:
- data:
isInIteration: false
sourceType: start
targetType: llm
id: 1727233123593-source-1732778846154-target
selected: false
source: '1727233123593'
sourceHandle: source
target: '1732778846154'
targetHandle: target
type: custom
zIndex: 0
- data:
isInIteration: false
isInLoop: false
sourceType: llm
targetType: code
id: 1732778846154-source-1748241838331-target
source: '1732778846154'
sourceHandle: source
target: '1748241838331'
targetHandle: target
type: custom
zIndex: 0
- data:
isInIteration: false
isInLoop: false
sourceType: code
targetType: end
id: 1748241838331-source-1727235861736-target
source: '1748241838331'
sourceHandle: source
target: '1727235861736'
targetHandle: target
type: custom
zIndex: 0
nodes:
- data:
desc: "{\n \"msg_info\": \"DeryckWhibley坚称他'不是骗子',前经理否认了性胁迫指控\",\n \"sys.files\"\
: [],\n \"sys.user_id\": \"sl_backend_twitter\"\n}"
selected: false
title: 开始
type: start
variables:
- label: msg_info
max_length: 2000
options: []
required: true
type: paragraph
variable: msg_info
height: 197
id: '1727233123593'
position:
x: -39.19210180980667
y: 274.1424650967822
positionAbsolute:
x: -39.19210180980667
y: 274.1424650967822
selected: false
sourcePosition: right
targetPosition: left
type: custom
width: 243
- data:
desc: ''
outputs:
- value_selector:
- '1748241838331'
- result
variable: result
selected: false
title: 结束
type: end
height: 89
id: '1727235861736'
position:
x: 1192.3926665576485
y: 328.2744664355823
positionAbsolute:
x: 1192.3926665576485
y: 328.2744664355823
selected: false
sourcePosition: right
targetPosition: left
type: custom
width: 243
- data:
context:
enabled: false
variable_selector: []
desc: ''
model:
completion_params:
top_p: 0.7
mode: chat
name: deepseek-v3-250324
provider: langgenius/volcengine_maas/volcengine_maas
prompt_template:
- id: ff06405d-dfb9-40ca-987a-c24fd3cfd6fb
role: system
text: '```xml
<instruction>
你是一个专业的数据筛选总结归纳者,请根据提供的文本内容,按照以下要求进行结构化整理:
1. 分类要求:
- 常见礼仪(细分:尊长爱幼/热情好客/社交礼仪)
- 着装习惯(细分:男性/女性/现代日常)
- 餐饮礼仪(细分:饮食特色/传统食物/节日饮食/待客习惯)
- 小费
- 宗教习惯(细分:宗教信仰/饮食/丧葬/节日)
- 总结提示(细分:尊重习俗/着装建议/饮食注意)
2. 格式规范:
- 每项内容不超过20字(细分条目允许完整句子)
- 输出严格采用JSON数组格式
- 禁止使用XML标签或任何非JSON格式符号
3. 操作步骤:
(1) 首先提取原文中与六大分类相关的所有信息
(2) 对每个大类下的细分条目进行语义压缩,确保核心信息完整,且细分只总结一条
(3) 将"常见礼仪"中的尊长爱幼、热情好客、社交礼仪分别总结为独立条目
(4) 将"着装习惯"中的男性、女性、现代日常分别总结为独立条目
(5) 将"餐饮礼仪"中的饮食特色、传统食物、节日饮食、待客习惯分别总结为独立条目
(6) 将"宗教习惯"中的宗教信仰、饮食、丧葬、节日分别总结为独立条目
(7) 将"总结提示"中的尊重习俗、着装建议、饮食注意分别总结为独立条目
(4) 特别注意宗教习惯中关于饮食禁忌的隐含信息
4. 注意事项:
- 当信息不明确时(如小费文化),标注"未提及"
- 传统服饰描述需保留关键特征词(如"缎纹布长外衣")
- 节日饮食要区分喜庆日与丧葬场合
- 除了小费分类以外其他分类中的每种细分条目只总结一条
</instruction>
<examples>
示例1:
输入:"在家庭聚会时年轻人需主动为长辈斟茶"
输出:[{"key":"常见礼仪","value":["尊长爱幼:年轻人主动为长辈斟茶"]}]
示例2:
输入:"男性传统头饰为羔羊皮制成的库拉帕克"
输出:[{"key":"着装习惯","value":["男性:头戴羔羊皮库拉帕克"]}]
示例3:
输入:"丧宴禁止饮酒但可食用抓饭"
输出:[{"key":"宗教习惯","value":["丧葬:丧宴禁酒允许食用抓饭"]},{"key":"餐饮礼仪","value":["节日饮食:丧葬场合食用抓饭"]}]
</examples>
<quality_control>
为确保输出质量,请核查:
1. 每个大类是否完整包含所有细分条目
2. 餐饮礼仪是否区分日常与节日场景
3. 宗教习惯是否包含信仰派别说明
4'
- id: 9f485eef-bc41-4f14-b8dd-9d507b6a8d3d
role: user
text: '{{#1727233123593.msg_info#}}'
selected: true
title: LLM
type: llm
variables: []
vision:
enabled: false
height: 89
id: '1732778846154'
position:
x: 350.4447530464781
y: 328.2744664355823
positionAbsolute:
x: 350.4447530464781
y: 328.2744664355823
selected: true
sourcePosition: right
targetPosition: left
type: custom
width: 243
- data:
code: "def main(text: str) -> dict:\n json_part = text.replace(\"```json\"\
, \"\").replace(\"```\", \"\").replace('\\\\n', '').strip().replace('\\\\\
','')\n data = json.loads(json_part)\n return {\n \"result\"\
: data,\n }"
code_language: python3
desc: ''
outputs:
result:
children: null
type: array[object]
selected: false
title: 代码执行
type: code
variables:
- value_selector:
- '1732778846154'
- text
variable: text
height: 53
id: '1748241838331'
position:
x: 734.3330920867545
y: 328.2744664355823
positionAbsolute:
x: 734.3330920867545
y: 328.2744664355823
selected: false
sourcePosition: right
targetPosition: left
type: custom
width: 243
viewport:
x: 260.89817163947
y: 285.1649665135767
zoom: 0.5370135153705135
Markdown is supported
0% or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment