Daily A-Share Market Snapshot

The Problem

After an A-share trading day, engineers and researchers often need a compact, shareable digest of index moves, sector and concept rankings, key company performance, and same-day news. Raw market JSON is fragmented, HK and A-share units differ, and news feeds easily mix old events with today's catalysts. Hand-writing HTML repeatedly introduces clipped charts, encoding issues, and inconsistent styling. This skill turns the workflow into a fixed pipeline: collect, validate, then generate a self-contained HTML report.

How It Works

• Fixed coverage: it defaults to seven indices such as SSE Composite, SZSE Component, STAR 50, CSI 300, CSI A500, CSI 1000, and Hang Seng Tech, plus 30 tracked HK and A-share companies.
• Structured market data: numeric data is collected through collect.py against Tencent-based market sources, with internal concurrency, per-response validate, and serial retries. It writes the report_data.json fields for indices, emotion, sectors, concepts, and company tables in one pass.
• News and commentary: the events module uses WebSearch on same-day authoritative finance sources, then filters for direct market impact or major company developments. The commentary module stays factual and avoids copying numbers already visible in charts or predicting price points.
• HTML generation: generate.py inlines ECharts and emits an offline-capable report with locked color, spacing, and chart behavior.

Boundaries

It is intended for objective daily, midday, or post-close market snapshots, not technical-analysis forecasting, moving-average or MACD signals, and buy/sell advice. If a data source is empty or a story is stale, the output should note the gap or exclude the item instead of inventing conclusions.