Data dependencies and execution prerequisites
It is recommended to use Wind wind-mcp-skill to obtain underlying data. Before execution, first check whether there is already a local wind-mcp-skill; if not, prompt the user to install:
# GitHub
npx skills add Wind-Information-Co-Ltd/wind-skills --skill wind-mcp-skill -g -y
# Gitee 镜像（国内）
npx skills add https://gitee.com/wind_info/wind-skills.git --skill wind-mcp-skill -g -y
Skill positioning
This skill is used to standardize the post-close review work for a group of watchlist stocks, helping users answer three things: who is strong and who is weak today, what is behind the strength and weakness, and whether it is still worth continuing to watch tomorrow. The goal is to provide a clearer observation list for the next trading day, rather than just summarizing price changes.
Execution process
Step 1: Review the day's performance
Check each stock's key performance for the day one by one, including:
- Price change, amplitude, trading volume changes, turnover changes
- Whether there was an intraday spike and pullback, a dip and rebound, or unusual movement near the close
- Compared with its sector and index, whether it led gains, followed gains, or clearly lagged behind
Step 2: Explain performance drivers
First determine which type of factor mainly drove today's performance:
- New news or announcements
- Sector rotation or market style switching
- Technical level trigger
- Capital games or emotional venting
- No obvious catalyst, only market fluctuation
Step 3: Identify the nature of strength and weakness
Divide strength and weakness into types with more decision-making significance, rather than only looking at gains and losses:
- Active strength: remains strong even after detaching from the sector
- Passive gain-following: clearly driven by the sector
- Turning weak under pressure: rebound is weak or structure is broken
- Noise fluctuation: no clear change in the main line
Step 4: Distill points to watch the next day
For each stock worth continuing to track, distill one most critical verification point for the next day, for example:
- Whether it can continue the strong performance with increased volume
- Whether it will fill the gap
- Whether it holds the newly formed key level
- Whether there is still continued event catalysis
Step 5: Update attention priority
Provide suggestions for adjusting attention levels the next day:
- Raise priority: a new main line, new strength, or new catalyst appeared today
- Maintain priority: the logic is unchanged, but there is no incremental signal for now
- Lower priority: structure is broken, catalyst falls through, or attention has clearly declined
Step 6: Output the post-close review summary
The final output should focus on "results + explanation + next-day observation" and should not show the step-by-step investigation process.
Output structure
Watchlist post-close review ({YYYY-MM-DD})
1. The most important review conclusions today
- Strongest observation target: {stock/reason}
- Weakest observation target: {stock/reason}
- The clue most worth carrying into tomorrow: {clue}
2. Review by stock
| Stock | Day's performance | Driver explanation | Nature of strength/weakness | Next-day observation point | Attention level |
|---|---|---|---|---|---|
| {Stock A} | {brief description} | {reason} | Active strength | {what to watch} | Raise |
| {Stock B} | {Brief description} | {Reason} | Under pressure and weakening | {What to watch} | Reduce |
III. Sector and Market Background
- {Which sectors have a clear impact on watchlist stock performance}
- {Whether the market style has changed}
IV. Tomorrow's Preparation
- Priority watchlist for tomorrow: {Stock list}
- Stocks that need confirmation: {Stock list}
- Stocks that can be temporarily removed from core focus: {Stock list}
Quality requirements
- The review must answer “why today turned out this way,” not just record the results.
- The conclusions should serve next-day preparation and avoid being written as a purely record-keeping daily report.
- Stocks with no incremental changes can be written briefly; do not spread effort evenly.
- Strength and weakness judgments must take sectors and the market environment into account; do not evaluate them in isolation from the background.
- If no clear driver can be found, write “no clear new driver for now” to avoid overexplaining.