stock

Name: stock
Author: MagicCube

489stars57forks

How to install stock?

magiccube/agentara/stock

$npx -y skills add MagicCube/agentara --skill stock

Installs into the current project.

Summary

Query stock quotes for US stocks, Hong Kong stocks, and A-shares. Use when the user asks about a stock price, market quote, or stock performance — including phrases like "查一下 BABA", "BABA 现在多少", "阿里巴巴股价", "茅台今天", "港股 9988", "美股行情", or any mention of a specific stock ticker or com

Files · 1

View on GitHub

SKILL.md

1	# Stock Quote Skill
2
3	Fetch latest stock price, change, and anomaly detection for US stocks, HK stocks, and A-shares.
4	Uses stooq.com as data source — reliable, no auth, bypasses system proxy automatically.
5	Data is T+1 for all markets (previous trading day close). This is expected and normal.
6
7	---
8
9	## Step 1 — Identify Stock & Market
10
11	Parse the user's query:
12
13	- A-share: 6-digit code (000001, 600519, 300750) or Chinese company name → market = `cn`
14	- HK stock: 4–5 digit code (9988, 00700, 9999) or `XXXX.HK` format → market = `hk`
15	- US stock: Latin ticker (BABA, AAPL, TSLA, NVDA) or `BABA.US` format → market = `us`
16
17	For well-known companies, infer the market:
18	- 阿里巴巴 → BABA (US) + 09988 (HK) — user usually means one, ask if unclear
19	- 腾讯 → 00700 (HK)
20	- 茅台/贵州茅台 → 600519 (A-share)
21	- 宁德时代 → 300750 (A-share)
22
23	For HK symbols, pass as-is (no zero-padding): `9988`, `700`. stooq handles it.
24
25	---
26
27	## Step 2 — Fetch Data
28
29	Use the unified `fetch_stock.py` script. Market codes: `hk`, `cn`, `us`.
30
31	```bash
32	# HK stock
33	python3 /Users/henry/.agentara/.claude/skills/stock/fetch_stock.py hk 9988
34
35	# A-share
36	python3 /Users/henry/.agentara/.claude/skills/stock/fetch_stock.py cn 600519
37
38	# US stock
39	python3 /Users/henry/.agentara/.claude/skills/stock/fetch_stock.py us BABA
40	```
41
42	Output is a JSON array of rows: `{date, open, close, high, low, vol, pct, chg}`.
43	If the output starts with `{"error":`, report the error to the user and stop.
44
45	Network notes: stooq.com is called with `trust_env=False` (bypasses Clash/system proxy).
46	No external Python deps beyond `requests` (stdlib-equivalent, always available).
47
48	Stooq symbol format:
49	- HK: `9988.hk` (no leading zeros)
50	- CN: `600519.cn`
51	- US: `baba.us` (lowercase)
52
53	---
54
55	## Step 3 — Analyze & Detect Anomalies
56
57	From the JSON output (up to 21 rows):
58
59	1. Latest row = most recent trading day data
60	2. Latest price = `close` of last row
61	3. Latest change% = `pct` of last row
62	4. Avg volatility = mean of `\|pct\|` of rows 2–21 (excluding last row)
63	5. Avg volume = mean of `vol` of rows 2–6 (5-day avg, excluding last)
64	6. Latest volume = `vol` of last row
65
66	Anomaly flags (report any that apply):
67	- `\|pct\| > 5%` → 🚨 大幅波动
68	- `\|pct\| > 2%` AND `\|pct\| > 2 × avg_volatility` → ⚠️ 异常波动（超出近期均值2倍）
69	- `latest_vol > 2 × avg_volume` → 📊 成交量异常放大（X倍）
70	- `latest_vol < 0.4 × avg_volume` → 🔇 成交量异常萎缩
71
72	If US stock `pct` is always near 0 except the last row, it's because the diff is computed on fetched data — use the last row's `pct` directly.
73
74	---
75
76	## Step 4 — Format Output
77
78	Keep it brief and mobile-friendly. Use list format, no tables.
79
80	Color + emoji rules (Chinese convention: 红涨绿跌):
81	- `pct > 0` → emoji 📈, wrap numbers in `<font color='red'>...</font>`
82	- `pct < 0` → emoji 📉, wrap numbers in `<font color='green'>...</font>`
83	- `pct == 0` → emoji ➡️, no color wrapper
84
85	Template (adapt to context):
86
87	```
88	{公司名} · {市场} {代码}
89
90	- 最新价：<font color='red/green'>{price} {货币}</font>
91	- 涨跌：{emoji} <font color='red/green'>{chg:+.2f} / {pct:+.2f}%</font>
92	- 最新交易日：{date}
93
94	{anomaly_section_if_any}
95	```
96
97	Anomaly section (only if flags exist):
98	```
99	⚠️ 异常提示
100	- [flag 1]
101	- [flag 2]
102	近5日均幅 {avg_vol:.2f}%，今日 <font color='red/green'>{pct:.2f}%</font>
103	```
104
105	If no anomaly: omit the anomaly section entirely. Keep the whole response under 10 lines.
106
107	---
108
109	## Step 5 — Currency
110
111	- A-share: ¥ (人民币)
112	- HK: HK$ (港元)
113	- US: $ (美元)
114
115	---
116
117	## Notes
118
119	- US data lags by 1 trading day (Sina source) — this is expected, tell the user if they ask
120	- If AKShare is not installed: `pip3 install akshare -q`
121	- If the symbol is wrong or not found, tell the user and suggest the correct format
122	- For A-share names, you can infer the 6-digit code from common knowledge; if uncertain, say so
123	- Do NOT show raw JSON or DataFrame output to the user — always parse and format it

People who liked this also viewed

6 picks

Type

stock

489stars57forks

View on GitHub

How to install stock?

magiccube/agentara/stock

$npx -y skills add MagicCube/agentara --skill stock

Installs into the current project.

Summary

Files · 1

View on GitHub

SKILL.md

1	# Stock Quote Skill
2
3	Fetch latest stock price, change, and anomaly detection for US stocks, HK stocks, and A-shares.
4	Uses stooq.com as data source — reliable, no auth, bypasses system proxy automatically.
5	Data is T+1 for all markets (previous trading day close). This is expected and normal.
6
7	---
8
9	## Step 1 — Identify Stock & Market
10
11	Parse the user's query:
12
13	- A-share: 6-digit code (000001, 600519, 300750) or Chinese company name → market = `cn`
14	- HK stock: 4–5 digit code (9988, 00700, 9999) or `XXXX.HK` format → market = `hk`
15	- US stock: Latin ticker (BABA, AAPL, TSLA, NVDA) or `BABA.US` format → market = `us`
16
17	For well-known companies, infer the market:
18	- 阿里巴巴 → BABA (US) + 09988 (HK) — user usually means one, ask if unclear
19	- 腾讯 → 00700 (HK)
20	- 茅台/贵州茅台 → 600519 (A-share)
21	- 宁德时代 → 300750 (A-share)
22
23	For HK symbols, pass as-is (no zero-padding): `9988`, `700`. stooq handles it.
24
25	---
26
27	## Step 2 — Fetch Data
28
29	Use the unified `fetch_stock.py` script. Market codes: `hk`, `cn`, `us`.
30
31	```bash
32	# HK stock
33	python3 /Users/henry/.agentara/.claude/skills/stock/fetch_stock.py hk 9988
34
35	# A-share
36	python3 /Users/henry/.agentara/.claude/skills/stock/fetch_stock.py cn 600519
37
38	# US stock
39	python3 /Users/henry/.agentara/.claude/skills/stock/fetch_stock.py us BABA
40	```
41
42	Output is a JSON array of rows: `{date, open, close, high, low, vol, pct, chg}`.
43	If the output starts with `{"error":`, report the error to the user and stop.
44
45	Network notes: stooq.com is called with `trust_env=False` (bypasses Clash/system proxy).
46	No external Python deps beyond `requests` (stdlib-equivalent, always available).
47
48	Stooq symbol format:
49	- HK: `9988.hk` (no leading zeros)
50	- CN: `600519.cn`
51	- US: `baba.us` (lowercase)
52
53	---
54
55	## Step 3 — Analyze & Detect Anomalies
56
57	From the JSON output (up to 21 rows):
58
59	1. Latest row = most recent trading day data
60	2. Latest price = `close` of last row
61	3. Latest change% = `pct` of last row
62	4. Avg volatility = mean of `\|pct\|` of rows 2–21 (excluding last row)
63	5. Avg volume = mean of `vol` of rows 2–6 (5-day avg, excluding last)
64	6. Latest volume = `vol` of last row
65
66	Anomaly flags (report any that apply):
67	- `\|pct\| > 5%` → 🚨 大幅波动
68	- `\|pct\| > 2%` AND `\|pct\| > 2 × avg_volatility` → ⚠️ 异常波动（超出近期均值2倍）
69	- `latest_vol > 2 × avg_volume` → 📊 成交量异常放大（X倍）
70	- `latest_vol < 0.4 × avg_volume` → 🔇 成交量异常萎缩
71
72	If US stock `pct` is always near 0 except the last row, it's because the diff is computed on fetched data — use the last row's `pct` directly.
73
74	---
75
76	## Step 4 — Format Output
77
78	Keep it brief and mobile-friendly. Use list format, no tables.
79
80	Color + emoji rules (Chinese convention: 红涨绿跌):
81	- `pct > 0` → emoji 📈, wrap numbers in `<font color='red'>...</font>`
82	- `pct < 0` → emoji 📉, wrap numbers in `<font color='green'>...</font>`
83	- `pct == 0` → emoji ➡️, no color wrapper
84
85	Template (adapt to context):
86
87	```
88	{公司名} · {市场} {代码}
89
90	- 最新价：<font color='red/green'>{price} {货币}</font>
91	- 涨跌：{emoji} <font color='red/green'>{chg:+.2f} / {pct:+.2f}%</font>
92	- 最新交易日：{date}
93
94	{anomaly_section_if_any}
95	```
96
97	Anomaly section (only if flags exist):
98	```
99	⚠️ 异常提示
100	- [flag 1]
101	- [flag 2]
102	近5日均幅 {avg_vol:.2f}%，今日 <font color='red/green'>{pct:.2f}%</font>
103	```
104
105	If no anomaly: omit the anomaly section entirely. Keep the whole response under 10 lines.
106
107	---
108
109	## Step 5 — Currency
110
111	- A-share: ¥ (人民币)
112	- HK: HK$ (港元)
113	- US: $ (美元)
114
115	---
116
117	## Notes
118
119	- US data lags by 1 trading day (Sina source) — this is expected, tell the user if they ask
120	- If AKShare is not installed: `pip3 install akshare -q`
121	- If the symbol is wrong or not found, tell the user and suggest the correct format
122	- For A-share names, you can infer the 6-digit code from common knowledge; if uncertain, say so
123	- Do NOT show raw JSON or DataFrame output to the user — always parse and format it

People who liked this also viewed

6 picks

Type