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 Taosheng777/BigA-stock-skills --skill stock-screenergit clone --depth 1 https://github.com/Taosheng777/BigA-stock-skillsWrote 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/taosheng777/biga-stock-skills/stock-screener)<a href="https://agentmods.dev/skills/taosheng777/biga-stock-skills/stock-screener"><img src="https://agentmods.dev/badge/skills/taosheng777/biga-stock-skills/stock-screener/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/taosheng777/biga-stock-skills/stock-screener"><img src="https://agentmods.dev/badge/skills/taosheng777/biga-stock-skills/stock-screener.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 | $0.00135 | $0.02084 |
| Opus 5 | $0.00068 | $0.01042 |
| Sonnet 5 | $0.00027 | $0.00417 |
| Haiku 4.5 | $0.00014 | $0.00208 |
Grade A, and why
stock-screener 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 12d 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.
How it starts
The opening of the file, as written. The whole thing — 86 lines — stays where its author put it; the contents beside it link to each section on GitHub.
问财选股（stock-screener）
本 SKILL.md 即运行时唯一执行依据（自包含）。
定位：A 股投研副驾驶的入口环节——只做"筛池子 + 数据化打分排序"，出名单后交给 stock-buddy 深度研判。与 stock-buddy 严格不重叠。
能力边界（铁律）
| ✅ 做（本 skill） | ❌ 不做（交给 stock-buddy） |
|---|---|
| 意图 → 问财问句（模板/自由口述） | 上涨逻辑、催化剂判断 |
| 翻页抓全池子（带上限） | 估值贵不贵 |
| 去重、排除已持仓 | 技术结构 / 套牢盘解读 |
| 透明加权打分、Top-N 排序 | 风险点分析 |
| 输出 HTML 名单 + 对话精简名单 | 买入/观察/放弃裁决、止损止盈、仓位 |
越界自检：用户若问"这只能不能买 / 估值贵不贵 / 怎么分析 / 帮我看看风险"，不作答分析，软提示一句：
这属于深度研判。对该标的说「用 stock-buddy 分析 600xxx」即可。
运行流程
第 1 步 · 判意图
"选股/筛股/筛池子/找票/强势股/放量突破/某条件选几只" → 进本 skill。 若是"分析我的持仓/某只怎么样/要不要买" → 不接，提示用 stock-buddy。
第 2 步 · 载 Key
先 source ~/.zshrc。IWENCAI_API_KEY 缺失则口述取 Key 指引（见 ~/.claude/skills/hithink-market-query/SKILL.md：iwencai.com/skillhub → 登录 → 安装方式-Agent用户-复制 Key），不要把 Key 打印出来。
第 3 步 · 取条件
- 命中模板名（放量突破/低位反转/业绩成长/主力异动）→ 用
--preset <名>。 - 否则把用户自由口述改写为标准问财问句（保意图、术语标准化、带基础排除如
非ST），并把问财专属打分列词缀进尾部（主力资金流向 归母净利润同比增长 净资产收益率）→ 用--query "<问句>"。 - ⚠️ 问财只下发问句里出现过的指标列（追加指标词不改筛选结果集，放心加）。行情类打分列（换手率/涨跌幅/市盈率/市净率）不必写：2026-07-15 起脚本内置腾讯批量 enrichment 确定性覆盖（零问财额度）；腾讯失败自动降级为问财返回列并在 JSON
enrich.note标记。 - 模板目录与写问句经验见
references/strategy-presets.md；不确定池子量级时，先用 hithink 的 cli.py 探code_count。
第 4 步 · 调脚本
脚本路径 = 本 SKILL.md 所在目录下的 scripts/screen.py（手动安装时即 ~/.claude/skills/stock-screener/scripts/screen.py，插件安装时在插件目录内，按本文件实际位置定位）。
source ~/.zshrc
python3 <本skill目录>/scripts/screen.py --preset 放量突破 --top 15
# 或自由口述：
python3 <本skill目录>/scripts/screen.py --query "今日主力净流入居前 换手率大于5% 非ST" --top 15
常用参数：--top N（默认 15）、--max-records（抓取上限，默认 100）、--no-exclude-holdings（不排除持仓）、--no-enrich（跳过腾讯行情覆盖，仅调试用）、--date。
脚本输出一段 JSON：fetched/returned/holdings_excluded/enrich{enriched,hit,note}/html_path/top[]。enrich.enriched=false 时在对话里说明"行情打分列降级为问财返回列"。
第 5 步 · 回结果
读脚本 JSON，对话内给：
- 精简名单：序号 / 代码 / 名称 / 综合分（+命中模板）；
- HTML 路径：
html_path（可排序表格，含分项贡献）； - 若
holdings_missing=true，说明"未找到持仓文件，未排除已持仓"； - 软提示交接：让用户自己挑 1–3 只（控成本）对其说「用 stock-buddy 分析」；
- 末尾固定一行非投顾声明（见下）。
What ships with it
8 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.
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.
- 12d ago First seen · 86 lines · 135 tokens per session scan A 758ebb92e528
stock-screener is a skill published in the GitHub repository Taosheng777/BigA-stock-skills (2 stars, last pushed 1mo ago), licensed MIT. It adds 135 tokens to every session and 2,084 once invoked, about $0.0007 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other skills, from other repositories
payments-intelligence-post-risk
Score a transaction for fraud risk by cross-checking the card BIN country against IP geolocation and billing country. Detects VPN, proxy, and TOR use.
payments-intelligence-post-validate
Validate and cross-check BIN, IBAN, and SWIFT identifiers in one call. Returns per-field results and a consistency check that flags country and bank mismatches.
bin-lookup-get-{bin}
Returns card metadata for the given 6–8 digit BIN prefix.
commodities-get-{commodity}
Returns the latest annual average price and up to 10 years of historical data for the requested commodity slug.
inflation-get-inflation
Returns the latest annual CPI inflation rate for a country plus the previous 10 years of historical data.
inflation-post-batch
Returns inflation data for up to 50 countries in a single request. Results are in the same order as the input. Countries with no data return found: false instead of failing the whole request. Billing: 1 credit per country (not per HTTP request).