Skills of A-share index data released by ft.tech.
v1.0.0A 股指数数据技能集(market.ft.tech)。覆盖单只指数详情、指数分页列表(排序/筛选)、指数 K 线(日/周/月/年线)、指数分钟级分时。用户询问某只指数行情、指数列表、指数 K 线或分时时使用。
Security Scan
OpenClaw
Benign
high confidencePurpose & Capability
Name/description match the implemented handlers. All handlers make GET requests to market.ft.tech endpoints described in the SKILL.md and return JSON; no unrelated services, binaries, or credentials are requested.
Instruction Scope
SKILL.md describes routing to sub-skills and how to invoke run.py. The runtime handlers only perform HTTP GET requests to the documented endpoints and print JSON. There are no instructions to read arbitrary files, environment variables, or to exfiltrate data to other domains.
Install Mechanism
No install spec or external downloads. This is an instruction/script bundle (Python scripts) that run locally; nothing is fetched or extracted during install.
Credentials
The skill declares no required env vars, credentials, or config paths. The code does not access environment variables or secrets; it only uses hardcoded BASE_URL and a fixed request header (X-Client-Name).
Persistence & Privilege
always is false and the skill does not modify other skills or system configuration. It contains Python scripts that will be executed when invoked, which is normal for a code-bearing skill.
Assessment
This skill is internally consistent: it only issues GET requests to https://market.ft.tech and returns the responses. Things to consider before installing: (1) the skill contains executable Python code stored with the skill bundle — installing it allows that code to run locally when the skill is invoked, so review the files (already included) if you need to be certain; (2) the skill makes network requests to market.ft.tech — ensure you trust that domain and its data/privacy posture; (3) there are no credential requests, but the skill will transmit whatever query parameters you pass (e.g., index codes) to the external service; (4) if you require stricter control, run the skill in a sandboxed environment or review/modify the handlers to restrict network access.Like a lobster shell, security has layers — review code before you run it.
latest
FT 指数数据 Skills
本 skill 是 FTShare-index-data 的统一路由入口。
根据用户问题,从下方「能力总览」或「询问方式与子 skill 对应表」匹配对应子 skill,然后通过 run.py 执行并解析响应。
所有接口均以
https://market.ft.tech为基础域名,请求头需携带X-Client-Name: ft-web(各子 skill 脚本已内置)。
调用方式(唯一规则)
run.py 与本文件(SKILL.md)位于同一目录。执行时:
- 取本文件的绝对路径,将末尾
/SKILL.md替换为/run.py,得到<RUN_PY>。 - 调用:
python <RUN_PY> <子skill名> [参数...]
# 示例(<RUN_PY> 为实际绝对路径)
python <RUN_PY> index-description-all
python <RUN_PY> index-detail --index 000001.XSHG
python <RUN_PY> index-list-paginated --order_by "change_rate desc" --page_size 20 --page_no 1
python <RUN_PY> index-ohlcs --index 000001.XSHG --span DAY1 --limit 50
python <RUN_PY> index-prices --index 000001.XSHG --since TODAY
python <RUN_PY> get-nth-trade-date --n 5
run.py内部通过__file__自定位,无论安装在何处都能正确找到各子 skill 的脚本。
指数 — 询问方式与子 skill 对应表
| 询问方式(用户常说的词) | 子 skill |
|---|---|
| 全部指数基础信息、指数列表(PB/PE)、有哪些指数、指数 简称/全称、市净率/市盈率 TTM | index-description-all |
| 某只 指数详情、上证指数行情、沪深300 点位/涨跌幅、指数名称/成交 | index-detail |
| 指数列表、全市场指数、按涨跌幅排序的指数、筛选某类指数 | index-list-paginated |
| 某只指数的 K 线、上证指数日线/周线/月线/年线、指数 开高低收、MA5/MA10/MA20 | index-ohlcs |
| 某只指数 分时、上证指数当日分时、指数 一分钟行情、多日分时走势 | index-prices |
| 前 N 个交易日、近 N 天交易日、往前推 N 个交易日(查近几天 K 线时先调此接口再转时间戳) | get-nth-trade-date |
能力总览
get-nth-trade-date:获取当前日期的前 N 个交易日。必填:--n(≥1)。查「近 N 天」K 线时先调本接口得到nth_trade_date,再按东八区转为毫秒时间戳用于 index-ohlcs 等。index-description-all:查询全部指数基础信息(symbol、全称、简称、pb、pe_ttm)。无需参数;GET /data/api/v1/market/data/index-description-all。index-detail:查询单只指数详情(名称、行情点位、成交、涨跌幅、多周期涨跌幅等)。必填:--index;可选--masks。若用户仅给名称,先通过index-description-all或index-list-paginated确认代码再查。index-list-paginated:指数分页列表,支持分页、排序、筛选。可选:--order_by/--ob、--filter、--masks、--page_size、--page_no。index-ohlcs:查询单只指数 OHLC K 线(开高低收、成交量、成交额),附带 MA5/MA10/MA20。必填:--index、--span(DAY1/WEEK1/MONTH1/YEAR1);可选--limit、--until_ts_ms。建议先完成名称到代码映射后再调用。index-prices:查询单只指数分钟级分时价格。必填:--index;时间范围二选一:--since(TODAY、FIVE_DAYS_AGO、TRADE_DAYS_AGO(n))或--since_ts_ms。建议先完成名称到代码映射后再调用。
使用流程
- 记录本文件绝对路径,将
/SKILL.md替换为/run.py得到<RUN_PY>。 - 理解用户意图,从「询问方式与子 skill 对应表」或「能力总览」匹配子 skill 名称。
- 若用户给的是指数名称/简称:先调用
index-description-all或index-list-paginated获取候选,确定标准代码(如000001.XSHG)。 - (可选)读取
sub-skills/<子skill名>/SKILL.md了解接口与参数。 - 执行:
python <RUN_PY> <子skill名> [参数...],获取 JSON 输出(详情/K 线/分时统一使用代码参数--index)。 - 解析并输出:以表格或要点形式展示给用户;若候选代码不唯一,先让用户确认再查询指标。
Comments
Loading comments...
