brainstorming
brainstorming
You MUST use this before any creative work - creating features, building components, adding functionality, modifying behavior, or planning anything new, in software or out of it (a talk, a business, a renovation).
hlzd-pipeline-vizB2B Skill 链路可视化 —— 消费各 Skill trace JSON,输出 8 KPI 仪表盘 + 漏斗图 + 评级分布 + 决策分布 + halt 警报 self-contained HTML(纯 stdlib + inline SVG,零依赖)。Use when 用户说'链路可视化'、'Pipeline 仪表盘'、'B2B 转化漏斗'、'Funnel Dashboard'、'pipeline metrics'、'trace aggregate'、'B2B analytics'。
Independent GitHub snapshot · Automated classification · Original documentation governs
把多个 Skill 的 trace JSON 累计成一份 self-contained HTML dashboard。
| KPI | 来源 |
|---|---|
| Traces | 输入 trace 数 |
| Halts | compliance/fraud 拦截数 |
| Follow-up actions due | followup-sequencer 输出 |
| Buyers evaluated | diligence 输出 |
| Quote total (USD) | quotation-gen 累计 |
| Avg margin | quotation-gen ratio |
| Knowledge entities | knowledge-graph 累计 |
| Graph nodes/edges | knowledge-graph 累计 |
import json
import lib
traces = [json.loads(p) for p in ["trace1.json", "trace2.json"]]
result = lib.run_pipeline(traces, title="HLZD Pipeline")
html = result["html_dashboard"]
open("dashboard.html", "w").write(html)
result["html_dashboard"] —— 直接浏览器打开result["aggregation"] —— 累计指标 JSON (for further processing)| 版本 | 说明 |
|---|---|
| 0.1.0 | 首版:8 KPI + 4 图表 + halt 警报 + SVG 自包含 |