A-Share Limit-Up Analysis with AI Agents
How short-term traders use an AI agent to automate A-share limit-up reviews: pull the limit-up list, reasons, consecutive boards and Dragon-Tiger institutional seats in one chain — no manual digging.

Tool names in this guide are for reference only and may have changed or been renamed. Please use the latest tools in the tool marketplace when implementing.
Key takeaways
- Ask an agent in plain language to pull a trading day's limit-up list and reasons at once
- Consecutive-board ladders and Dragon-Tiger institutional seats chain into the same flow
- The review report is template-driven, so you can run the same pass after every close
Tools used
Why limit-up reviews are so tedious
For short-term and hot-money traders, the post-close limit-up review is a daily ritual: which stocks hit the limit up, why, how many consecutive boards, and who was buying on the Dragon-Tiger list. The trouble is that this information is scattered — the limit-up list sits in one place, each reason has to be checked stock by stock, the board count is yours to tally, and the Dragon-Tiger seats live in yet another table. Stitching it together by hand eats an entire evening, and things slip through.
If you're searching for A-share limit-up analysis, what you usually want isn't another watch-list screen but one flow that chains "limit-up → reason → boards → Dragon-Tiger" automatically and produces a report every day.
How this workflow breaks down
Picture a full limit-up review as one closed loop:
- Give the agent a review date.
- Pull that day's limit-up list and reasons.
- Check limit prices and identify the consecutive-board ladder.
- Query the Dragon-Tiger list and institutional seats to see who was trading.
- Have the agent compile it into a review report.
You just state the date and the angles you care about; tool selection and stitching are the agent's job.
How each step works
Step 1: Give the review date
Say it in plain language: use ae to review the A-share limit-ups for a given trading day. Pass the date as a parameter, and every following step revolves around that session.
Step 2: Pull the limit-up list and reasons
A limit-up board tool (such as ae_tushare_kpl_list) returns the stocks that hit the limit up that day along with the reason / theme tags. This answers "who limited up today and why" — the reason tags are the most time-consuming part of a review, and you get them in one call.
Step 3: Check limit prices and boards
A limit-price tool (such as ae_tushare_stk_limit) gives each stock's upper and lower limit prices. Combined with the limit-up list, the agent can tell first boards from consecutive ones and lay out the board ladder.
Step 4: Query the Dragon-Tiger list and seats
A Dragon-Tiger tool (such as ae_tushare_top_list) returns the buy/sell amounts for listed stocks, and an institutional-seat tool (such as ae_tushare_top_inst) adds the specific seat breakdown. This answers "who was buying, who was selling, and whether institutions stepped in."
| Step | Tool category | Output |
|---|---|---|
| Limit-up list + reasons | Limit-up board tool (ae_tushare_kpl_list) |
Limit-up stocks and theme reasons |
| Limit prices / boards | Limit-price tool (ae_tushare_stk_limit) |
Consecutive-board ladder |
| Dragon-Tiger list | Dragon-Tiger tool (ae_tushare_top_list) |
Listed buy/sell amounts |
| Institutional seats | Seat tool (ae_tushare_top_inst) |
Specific seat breakdown |
Step 5: Produce the review report
With all four steps in hand, have the agent compile to a fixed template: use ae to turn the above into a limit-up review report, sorted by board count, flagging stocks with institutional seats. The result is a structured report you can archive directly.
Manual review vs. agent-automated review
| Manual review | Agent-automated review | |
|---|---|---|
| Limit-up reasons | Checked stock by stock | Pulled in one call |
| Board ladder | Tallied and ranked by hand | Sorted automatically |
| Dragon-Tiger seats | A separate list to cross-check | Pulled in the same chain |
| Report | Hand-assembled doc | Template-driven, auto-written |
| Daily effort | A whole evening | One instruction |
FAQ
Q: Can I see the limit-up reasons? Yes. The limit-up board tool returns reason / theme tags for each limit-up stock — the biggest time saver in a review.
Q: Can I see Dragon-Tiger institutional seats? Yes. The Dragon-Tiger tool gives listed buy/sell amounts, and the seat tool adds the specific institutional breakdown, so who's trading is clear at a glance.
Q: Can it run on a schedule? Yes. Fix this set of instructions as a daily task and let the agent run the same flow after each close, producing the report on schedule.
How to get started
- Sign up — new users get 1,000 free credits.
- Connect in-app (WorkBuddy / Trae) or add an MCP config line for Claude Code / Cursor / VSCode.
- Give instructions in natural language and let the agent do the rest.
