feat(pipelines): add 10 industry-specific pipeline templates

Education (3): research-to-quiz, student-analysis, lesson-plan
Healthcare (3): policy-compliance, meeting-minutes, data-report
Design Shantou (4): trend-to-design, competitor-research,
  client-communication, supply-chain-collect
This commit is contained in:
iven
2026-04-01 23:43:45 +08:00
parent 17a2501808
commit c6bd4aea27
10 changed files with 2318 additions and 0 deletions

View File

@@ -0,0 +1,205 @@
# ZCLAW Pipeline - 医疗数据报告生成
# 输入科室运营数据,自动生成结构化运营分析报告
apiVersion: zclaw/v1
kind: Pipeline
metadata:
name: healthcare-data-report
displayName: 医疗数据报告生成
category: healthcare
industry: healthcare
description: 输入科室运营数据(门诊量、住院、手术、费用等),自动生成运营分析报告
tags:
- 医疗
- 数据报告
- 运营分析
- 行政管理
icon: 📈
author: ZCLAW
version: 1.0.0
spec:
inputs:
- name: data_content
type: text
required: true
label: 运营数据
placeholder: |
粘贴科室运营数据,支持格式:
指标名称,数值 或自由文本描述
例如:门诊量 1250人次同比增长8%
- name: report_period
type: select
required: false
label: 报告周期
default: 月报
options:
- 周报
- 月报
- 季报
- 半年报
- 年报
- name: department
type: string
required: false
label: 科室
placeholder: 例如:内科、外科、急诊科
- name: focus_areas
type: multi-select
required: false
label: 分析重点
default:
- volume_analysis
- quality_metrics
- financial
options:
- volume_analysis
- quality_metrics
- financial
- patient_satisfaction
- resource_utilization
steps:
# Step 1: 数据解析
- id: parse_data
description: 解析原始数据为结构化指标
action:
type: llm_generate
template: |
解析以下医疗运营数据为结构化指标:
报告周期: {{report_period}}
科室: {{department}}
原始数据:
```
{{data_content}}
```
请生成 JSON 格式:
{
"metrics": {
"outpatient": {"value": 0, "unit": "人次", "yoy_change": "同比变化"},
"inpatient": {"value": 0, "unit": "人次", "yoy_change": ""},
"surgery": {"value": 0, "unit": "台次", "yoy_change": ""},
"bed_occupancy": {"value": "0%", "unit": "", "yoy_change": ""},
"average_stay": {"value": "0天", "unit": "", "yoy_change": ""},
"revenue": {"value": "0万", "unit": "", "yoy_change": ""}
},
"data_quality": "数据完整/部分缺失",
"missing_fields": ["缺失指标"]
}
input:
data_content: ${inputs.data_content}
report_period: ${inputs.report_period}
department: ${inputs.department}
json_mode: true
temperature: 0.3
max_tokens: 2000
# Step 2: 同比环比分析
- id: trend_analysis
description: 同比环比趋势分析
action:
type: llm_generate
template: |
对以下医疗指标进行趋势分析:
科室: ${inputs.department}
报告周期: ${inputs.report_period}
结构化数据: ${steps.parse_data.output}
请生成 JSON 格式:
{
"trends": [
{
"metric": "指标名称",
"current": "当前值",
"previous": "上期值",
"change_rate": "变化率",
"trend": "上升/下降/持平",
"assessment": "正常/需关注/预警"
}
],
"highlights": ["亮点1", "亮点2"],
"concerns": ["关注点1", "关注点2"],
"overall_assessment": "整体评估"
}
input:
department: ${inputs.department}
report_period: ${inputs.report_period}
data: ${steps.parse_data.output}
json_mode: true
temperature: 0.4
max_tokens: 2500
# Step 3: 质量指标分析
- id: quality_analysis
description: 医疗质量指标分析
action:
type: llm_generate
template: |
基于运营数据进行医疗质量分析:
科室: ${inputs.department}
运营数据: ${steps.parse_data.output}
趋势分析: ${steps.trend_analysis.output}
请生成 JSON 格式:
{
"quality_indicators": [
{"indicator": "质量指标", "value": "当前值", "benchmark": "参考标准", "status": "达标/未达标"}
],
"safety_events": "安全事件统计",
"improvement_areas": ["改进方向1", "改进方向2"],
"recommendations": ["质量改进建议1", "质量改进建议2"]
}
input:
department: ${inputs.department}
data: ${steps.parse_data.output}
trends: ${steps.trend_analysis.output}
json_mode: true
temperature: 0.4
max_tokens: 2000
# Step 4: 生成完整报告
- id: generate_report
description: 生成完整运营分析报告
action:
type: llm_generate
template: |
生成完整的医疗运营分析报告:
科室: ${inputs.department}
周期: ${inputs.report_period}
请生成 JSON 格式报告:
{
"title": "报告标题",
"executive_summary": "200字摘要",
"data_overview": "数据概览",
"key_findings": ["关键发现1", "关键发现2", "关键发现3"],
"action_items": ["行动建议1", "行动建议2"],
"preview_text": "300字报告预览"
}
input:
department: ${inputs.department}
report_period: ${inputs.report_period}
data: ${steps.parse_data.output}
trends: ${steps.trend_analysis.output}
quality: ${steps.quality_analysis.output}
json_mode: true
temperature: 0.5
max_tokens: 2000
outputs:
parsed_data: ${steps.parse_data.output}
trends: ${steps.trend_analysis.output}
quality: ${steps.quality_analysis.output}
report: ${steps.generate_report.output}
on_error: stop
timeout_secs: 240