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 cyijun/china-financial-services --skill a-share-research-evidencegit clone --depth 1 https://github.com/cyijun/china-financial-servicesWrote 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/cyijun/china-financial-services/a-share-research-evidence)<a href="https://agentmods.dev/skills/cyijun/china-financial-services/a-share-research-evidence"><img src="https://agentmods.dev/badge/skills/cyijun/china-financial-services/a-share-research-evidence/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/cyijun/china-financial-services/a-share-research-evidence"><img src="https://agentmods.dev/badge/skills/cyijun/china-financial-services/a-share-research-evidence.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 | $0.00078 | $0.00792 |
| Opus 5 | $0.00039 | $0.00396 |
| Sonnet 5 | $0.00016 | $0.00158 |
| Haiku 4.5 | $0.00008 | $0.00079 |
Grade A, and why
a-share-research-evidence 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.
What it actually says
A股研究证据闸门
先建立可复核证据，再允许下游分析使用。
工作流
- 写明研究对象、研究问题、
as_of时刻、市场时区和允许的数据延迟。 - 建立证据台账：每条记录包含原始来源、文档日期、实际披露时刻、报告期、抓取时刻、口径、单位、币种、复权方式和定位信息。
- 关键数字优先回到交易所、巨潮资讯或公司法定披露。Tushare和AKShare属于结构化便利层，不能覆盖原始披露。
- 结构化取数使用
china-market-data。历史研究传入as_of和require_pit=true；若主源失败且备用源缺少历史可得时点，必须停止。 - 财务数据只能在
ann_date、f_ann_date或可核验披露时刻之后可见；修订数据保留原版本与新版本。 - 处理冲突时先检查合并范围、单季/累计、TTM、币种、复权、税前/税后、法定/调整后和发布时间。无法消解时并列记录，不平均。
- 输出可用、冲突、缺失和禁止使用四类证据，以及下游可安全引用的字段。
原文与定位清单
将证据记录写成JSON列表后运行scripts/build_evidence_manifest.py input.json --output evidence-manifest.json。每条至少含evidence_id、来源类型、HTTPS原文URL、文档日、带时区的实际披露时刻、研究as_of、页码/章节/表格/段落定位和支撑主张。已有本地原文可传local_path并计算内容SHA-256；显式传--download-dir时脚本才会只读下载HTTPS原文、限制单文件体积并落盘哈希。截止日后的证据会标为forbidden_future并使清单blocked。
下载成功只证明取得了某个字节版本，不证明来源权威、披露时间或内容解释正确；交易所/巨潮链接跳转后仍要核对最终URL、公告编号和文档内标题。
硬约束
- 搜索摘要不能代替原文；二手来源不足以单独支撑高影响结论。
- 报告期不是可得时点；披露计划也不是实际披露时刻。
- 不确定来源、单位或时间的数字标为
unverified，不用于精确计算。 - 研究截止日之后的信息只能作为事后验证，不能污染当时视角。
- 只做只读研究，不登录交易账户，不修改自选或持仓。
输出契约
先给证据状态ready、partial或blocked，随后提供研究时点、来源清单、证据台账、冲突处理、缺口和干净字段。
需要详细规则时读取references/source-and-pit-rules.md。
What ships with it
2 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 · 39 lines · 78 tokens per session scan A 575180c3db5d
a-share-research-evidence is a skill published in the GitHub repository cyijun/china-financial-services (19 stars, last pushed 18d ago), licensed Apache-2.0. It adds 78 tokens to every session and 792 once invoked, about $0.0004 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 2026-08-30.
Other skills, from other repositories
hithink-finance
A routing guide for accessing Chinese A-share financial data, including prices, company reports, valuations, funds, indices, sectors, and local data storage.
hithink-finance-fund
A command-line guide for querying fund information, including profiles, managers, holdings, prices, returns, financial data, news, and exchange-traded fund snapshots. A command-line tool is a program controlled by typed terminal commands.
hithink-finance-data
A local data-management skill for the HiThink Finance command-line tool and its DuckDB database. DuckDB is a database stored in a local file.
hithink-finance-market
A command-line tool entry for retrieving ordinary Chinese A-share market data, including snapshots, historical price bars, trading calendars, adjustment factors, and company actions.
hithink-finance-special-data
A command-line tool entry for retrieving special Chinese market lists and event data, such as limit-up stocks, limit-down stocks, unusual moves, hot stocks, and Dragon-Tiger records.
hithink-finance-futures
A command-line data source for public futures-market information, including contracts, positions, warehouse receipts, basis, trading schedules, and price charts. Futures are agreements to buy or sell an asset at a set future date.