Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add shouldnotappearcalm/a-share-skill --skill a-share-strategy-mainboard-multi-swing-defensivegit clone --depth 1 https://github.com/shouldnotappearcalm/a-share-skillWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/shouldnotappearcalm/a-share-skill/a-share-strategy-mainboard-multi-swing-defensive)<a href="https://agentmods.dev/skills/shouldnotappearcalm/a-share-skill/a-share-strategy-mainboard-multi-swing-defensive"><img src="https://agentmods.dev/badge/skills/shouldnotappearcalm/a-share-skill/a-share-strategy-mainboard-multi-swing-defensive/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/shouldnotappearcalm/a-share-skill/a-share-strategy-mainboard-multi-swing-defensive"><img src="https://agentmods.dev/badge/skills/shouldnotappearcalm/a-share-skill/a-share-strategy-mainboard-multi-swing-defensive.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00103 | $0.01804 |
| Opus 5.5 | $0.00041 | $0.00722 |
| Sonnet 5.5 | $0.00021 | $0.00361 |
| Haiku 4.5 | $0.00010 | $0.00180 |
Grade A, and why
a-share-strategy-mainboard-multi-swing-defensive scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 17d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
This is a copy
95% identical to a-share-strategy-allmarket-multi-swing-defensive — 10 lines differ, which has more behind it and is treated as the original. This page carries a canonical link to it rather than competing with it.
How it starts
The opening of the file, as written. The whole thing — 108 lines — stays where its author put it; the contents beside it link to each section on GitHub.
主板多波段防御型 · 决策信号
本 skill 核心做三件事：选股范围、买入侧信号、卖出侧信号；并可选使用 realtime_quotes.py 对指定代码批量拉取现价类快照（基于 MarketDataProvider.get_quote，腾讯报价 + 分钟 K 聚合）。
输出的是决策参考（结构化列表或 JSON），不包含历史回测撮合、不包含自动报单；真实下单请用券商或另接 a-share-paper-trading 等执行通道。
策略信号定位为条件有效：在市场结构与交易成本变化下，信号强度可能衰减，需持续复核。
能力边界
| 做 | 不做 |
|---|---|
| 从主板高流动性股票中取前 N 只构成当日股票池 | 分钟级回测、混合回测、收益曲线 |
用日线 trend_pullback 标出入场/离场条件 |
保证收益或替代投顾 |
| 给出两类「买入参考」列表（见下） | 直接向交易所或模拟盘服务下单 |
| 可选：读取持仓列表文件，标出「策略离场」标的 | 替你保存实盘持仓（除非你自建文件） |
批量拉取多只股票现价、涨跌幅、涨跌停参考价等（realtime_quotes.py） |
替代 a-share-data 的全市场实时板块、资金流等重接口 |
策略逻辑（与参数）
默认参数见 scripts/strategy_lab/strategy_params.py（均线快慢、回踩幅度、RSI 区间与离场 RSI 等）。
信号计算在 scripts/strategy_lab/strategies.py 的 trend_pullback。
脚本会额外应用两层过滤：
- 成本过滤：按
--roundtrip-cost-bps估算往返成本，只有edge_after_cost > 0的候选才进入最终买入列表。 - 鲁棒性过滤：在
ROBUSTNESS_PARAM_GRID参数邻域内计算entry_consensus_ratio，低于--entry-consensus-min的候选会被过滤。
买入参考（两组，请区分语义）：
from_previous_day_close：上一根已收盘日线满足entry。与「前一日收盘后出信号、当日再执行」的习惯一致，更贴近事前计划。from_last_close：最新一根日线也满足entry，偏形态展示；若与上一日重复，请避免重复计数。
列表内按策略内 score（均线强弱） 降序；默认每种买入列表最多保留 5 只（strategy_params.MAX_BUY_CANDIDATES），与「同时关注仓位不宜过多」一致。需要更多或全部时加 --max-buys 0 表示不截断，或 --max-buys 10 等。
卖出参考：
对 --holdings 文件中的代码，若最新一根日线满足 exit（破慢线或 RSI 过高等规则内条件），则列入卖出参考。文件格式：一行一只代码，可含注释行（# 开头）。
风控参考（仅文档与 JSON 字段）：
REFERENCE_INTRADAY_STOP_PCT 表示历史上与策略文档一致的日内止损比例参考，本脚本不替你监控盘中止损，需自行在下单软件中设置。
环境与依赖
pip install akshare pandas numpy requests
运行
SKILL_DIR="<本 skill 绝对路径>"
python3 "$SKILL_DIR/scripts/daily_decisions.py" --json
常用参数：
--top-n：股票池大小，默认 120--max-buys：买入侧列表在排序后最多保留几条，默认 5；传0不截断--holdings：持仓代码文件路径--workers：拉日线并发数--json：输出一份 JSON，便于程序消费（JSON 内含截断前数量*_total）--roundtrip-cost-bps：往返交易成本估计（单位 bps），默认 45--entry-consensus-min：参数邻域入场一致性阈值，默认 0.67--disable-robust-check：关闭参数邻域一致性过滤（仅在调试时使用）
What ships with it
9 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
- scripts/daily_decisions.py 14 KB runs code
- scripts/paper_trading/__init__.py 0 B runs code
- scripts/paper_trading/market_data.py 30 KB runs code
- scripts/realtime_quotes.py 5.8 KB runs code
- scripts/strategy_lab/__init__.py 0 B runs code
- scripts/strategy_lab/indicators.py 1.4 KB runs code
- scripts/strategy_lab/strategies.py 9.8 KB runs code
- scripts/strategy_lab/strategy_params.py 895 B runs code
- scripts/test_market_universe_filters.py 1.7 KB runs code
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 17d ago First seen · 108 lines · 103 tokens per session scan A 301cceda3701
a-share-strategy-mainboard-multi-swing-defensive is a skill published in the GitHub repository shouldnotappearcalm/a-share-skill (247 stars, last pushed 5d ago), licensed MIT. It adds 103 tokens to every session and 1,804 once invoked, about $0.0004 per session on Opus 5.5. A static security scan graded it A with 0 findings. It is 95% identical to a-share-strategy-allmarket-multi-swing-defensive, differing in 10 lines, and is treated as a copy.
Other skills, from other repositories
china-dcf-model
A documented discounted-cash-flow model for valuing non-financial A-share companies. A discounted-cash-flow model estimates today’s value from expected future cash flows and makes its assumptions, valuation date, and adjustments explicit.
pptx-author
Produce a .pptx file on disk for headless Chinese A-share market research, using python-pptx with Chinese font support. Use when no controllable Office session is available.
industry-etf-research
A research workflow for studying Chinese A-share industry ETFs and the indexes they track. It compares index rules, holdings, company fundamentals, industry conditions, valuations and market evidence.
china-market-data
A routing and quality-check process for obtaining Chinese stock-market and company-financial data from sources such as Tushare Pro and AKShare.
a-share-factor-validation
A research checklist and toolkit for testing stock-selection factors and backtests in China's A-share market, the mainland Chinese stock market. It checks whether results avoid future-data leakage and account for trading rules, costs, and statistical uncertainty.
china-market-researcher
A research workflow for analysing Chinese-listed stocks by industry or investment theme, including companies, competitors, ETFs, and market data. A-share means shares listed on mainland Chinese stock exchanges.