-
Notifications
You must be signed in to change notification settings - Fork 0
Expand file tree
/
Copy pathnode-skills.yaml
More file actions
87 lines (81 loc) · 3.28 KB
/
node-skills.yaml
File metadata and controls
87 lines (81 loc) · 3.28 KB
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
# node-skills.yaml
# HADS 心理评估 Skill 节点定义与编排
nodes:
# --- 核心主节点 (Master Node) ---
# 对应 SKILL.md 定义的完整业务逻辑
hads_assessment_master:
type: flow_master
description: "协调 HADS 问卷发布、结果解析及自优化闭环的总体流程"
next:
- call: hads_publisher
map_input: { patient_id: "{{input.patient_id}}" }
# --- 子节点 1: 问卷发布 (Publisher) ---
hads_publisher:
type: python_skill
cwd: ~/.claude/skills/skill-hads-assessment
entrypoint:
command: python
args: ["src/publisher.py", "--patient-id", "{{input.patient_id}}"]
output_schema:
type: object
properties:
questionnaire_html: { type: string }
# 发布后等待患者填写,逻辑上由 Agent 或系统触发下一步解析
next:
- call: hads_result_parser
map_input: { patient_id: "{{input.patient_id}}" }
# --- 子节点 2: 结果解析与智能解读 (Parser & Interpreter) ---
# 此节点集成了 Interpretation Agent 的提示词生成能力
hads_result_parser:
type: python_skill
cwd: ~/.claude/skills/skill-hads-assessment
entrypoint:
command: python
args: ["src/result_parser.py", "results/json/hads_result_{{input.patient_id}}.json"]
output_schema:
type: object
properties:
anxiety_score: { type: integer }
depression_score: { type: integer }
llm_prompt: { type: string } # Interpretation Agent 生成的深度解读指令
next:
- call: hads_self_optimizer
map_input: {}
# --- 子节点 3: 自优化反思 (Self-Optimization) ---
# 此节点集成了 Reflection Agent 对四级日志的分析能力
hads_self_optimizer:
type: python_skill
cwd: ~/.claude/skills/skill-hads-assessment
entrypoint:
command: python
args: ["src/optimizer.py"]
output_schema:
type: object
properties:
suggested_changes: { type: array }
needs_human_approval: { type: boolean }
# --- 子节点 4: 内容生成器 (Content Generator) ---
# 为患者生成主题化的静态 HTML 康复指导页面(如正念、营养、运动)
# 集成 Bilibili 精选视频,采用治愈系 UI 设计
hads_content_generator:
type: python_skill
cwd: ~/.claude/skills/skill-hads-assessment
entrypoint:
command: python
args: ["src/content_generator.py", "--theme", "{{input.theme}}", "--patient-id", "{{input.patient_id}}", "--output-prompt-only"]
output_schema:
type: object
properties:
llm_prompt: { type: string } # 用于 LLM 生成 HTML 的完整提示词
theme: { type: string }
description: >
生成用于 LLM 的 HTML 提示词,指导生成包含精选视频的治愈系静态页面。
实际 HTML 生成需要调用支持 Web 搜索的 LLM(如 Claude + MCP),
生成后可通过 EdgeOne MCP 发布到公网供患者访问。
flows:
# HADS 评估完整生命周期流
- name: hads_complete_assessment_lifecycle
entry: hads_assessment_master
description: >
从发布针对特定患者的 HADS 问卷开始,采集数据后进行自动化计分与智能临床解读建议生成,
最终通过分析执行日志输出系统优化提案,完成受控的自进化闭环。