Ready to supercharge your AI agents?

Install any skill in seconds. Free to start, no credit card required.

A\u80A1\u4E2A\u80A1\u6DF1\u5EA6\u7814\u7A76/\u9996\u6B21\u8986\u76D6\u62A5\u544A · aifinlab | DecimalAI

aifinlab/a-share-initiation

NewSkillScoreMaintained

A股个股深度研究/首次覆盖报告/个股全面研究分析。当用户说"深度研究"、"深度分析"、"全面分析XX"、"首次覆盖"、"initiation"、"initiating coverage"、"帮我研究一下XX"、"XX值不值得买"、"给我出一份XX的研究报告"、"个股深度"、"深度报告"、"投资价值分析"时触发。MUST USE when user asks for deep research, initiating coverage, comprehensive stock analysis, or a full research report on a specific stock. 整合公司概况、行业地位、财务分析、估值（DCF+可比）、投资论点，生成全面的个股研究报告。通过 cn-stock-data 获取全量数据，结合 web 搜索获取公司信息、行业背景、分析师观点。支持券商深度报告风格（formal，15-30页）和个人研究笔记风格（brief，3-5页）。

financeby aifinlab

Links this skill to your agents and keeps it current with the author’s updates.

Automatic

Finance & Accounting pack

1 · One-time — install the CLI (docs ↗)

pip install decimalai

2 · Get the skill — writes .claude/skills/aifinlab-a-share-initiation/SKILL.md

decimalai skills pull aifinlab-a-share-initiation --out .claude/skills/

No CLI? Get the same file — curl · raw file

curl -fsSL --create-dirs -o .claude/skills/aifinlab-a-share-initiation/SKILL.md https://app.decimal.ai/s/aifinlab-a-share-initiation/SKILL.md

Raw file: /s/aifinlab-a-share-initiation/SKILL.md ↗ · machine-readable summary at /s/aifinlab-a-share-initiation.json ↗

In productionGlobal

how this skill behaves on real traffic · last 30 days

No production data for this skill yet — usage numbers appear here once agents start running it.

Anonymized aggregate across every org using this skill — proof from real production use, not a demo.

Does this skill help? · Benchmark: with vs. withoutsign in to run this on your agents

24 test cases from the publisher’s eval.yaml · each run twice on gemini-3.6-flash — with the skill injected, then without · last run 2d ago

—+175% tokens0% turns21% of whole cases pass with skill (5/24)

Test case	Without → With	Effect	Δ tokens	Δ turns
case-16	✗→✓	▲ Improved	158%	0%
case-18	✗→✓	▲ Improved	91%	0%
case-20	✓→✗	▼ Worse	49%	0%
case-21	✓→✗	▼ Worse	53%	0%
case-22	✓→✗	▼ Worse	23%	0%

Showing 5 of 24 cases

SkillSafety

✓ 3 of 3 passed

Code scan15 of 15 checks clean

0 critical · 0 warning · 0 info — scanner v4 — 8/3/2026

live_secretremote_code_executionreverse_shelldata_exfiltrationssrf_metadataagent_snoopinganti_refusaldestructive_commandsobfuscated_unicodeunicode_homoglyphopaque_blobsuspicious_urltool_overreachtrigger_abuseprompt_injection_phrasing

IntentSafe

LLM judge · does the skill do what it claims

ContentClean

LLM judge · policy & harmful-content review

Source

Imported

aifinlab / finclaw

skills/a-share-initiation/SKILL.md

View on GitHub ↗

License Apache-2.0Imported 8/3/2026Version v1Bundled 1 file

Skill contents

references/

SKILL.md

A\u80A1\u4E2A\u80A1\u6DF1\u5EA6\u7814\u7A76/\u9996\u6B21\u8986\u76D6\u62A5\u544A

\u6570\u636E\u6E90

bash
SCRIPTS="$SKILLS_ROOT/cn-stock-data/scripts"

# \u5B8C\u6574\u8D22\u52A1\u6307\u6807\uFF08\u591A\u671F\uFF09
python "$SCRIPTS/cn_stock_data.py" finance --code [CODE]

# \u5B9E\u65F6\u884C\u60C5
python "$SCRIPTS/cn_stock_data.py" quote --code [CODE]

# \u8FD1 2 \u5E74\u65E5\u7EBF
python "$SCRIPTS/cn_stock_data.py" kline --code [CODE] --freq daily --start [2\u5E74\u524D\u65E5\u671F]

# \u8D44\u91D1\u6D41\u5411
python "$SCRIPTS/cn_stock_data.py" fund_flow --code [CODE] --days 30

# \u5317\u5411\u8D44\u91D1
python "$SCRIPTS/cn_stock_data.py" north_flow

# \u53EF\u6BD4\u516C\u53F8\u884C\u60C5
python "$SCRIPTS/cn_stock_data.py" quote --code [COMP1],[COMP2],[COMP3]

# \u53EF\u6BD4\u516C\u53F8\u8D22\u52A1
python "$SCRIPTS/cn_stock_data.py" finance --code [COMP]

\u8865\u5145 web \u641C\u7D22\uFF1A\u516C\u53F8\u5B98\u7F51\u3001\u5E74\u62A5\u6458\u8981\u3001\u7BA1\u7406\u5C42\u4ECB\u7ECD\u3001\u884C\u4E1A\u62A5\u544A\u3001\u5206\u6790\u5E08\u89C2\u70B9\u3001\u8FD1\u671F\u516C\u544A\u3002

Workflow (7 phases)

Phase 1: \u516C\u53F8\u6982\u51B5

Web \u641C\u7D22\uFF1A\u516C\u53F8\u4E3B\u8425\u4E1A\u52A1\u3001\u53D1\u5C55\u5386\u7A0B\u3001\u80A1\u6743\u7ED3\u6784\u3001\u7BA1\u7406\u5C42\u80CC\u666F
\u4E1A\u52A1\u62C6\u5206\uFF1A\u5404\u4E1A\u52A1\u677F\u5757\u6536\u5165\u5360\u6BD4\u3001\u589E\u901F
\u884C\u4E1A\u5B9A\u4F4D\uFF1A\u6240\u5C5E\u884C\u4E1A\u3001\u5E02\u573A\u5730\u4F4D\u3001\u5E02\u5360\u7387
\u6838\u5FC3\u7ADE\u4E89\u529B\uFF1A\u54C1\u724C/\u6280\u672F/\u6E20\u9053/\u89C4\u6A21/\u724C\u7167\u7B49\u58C1\u5792

Phase 2: \u884C\u4E1A\u5206\u6790\uFF08\u7CBE\u7B80\u7248 a-share-sector\uFF09

\u884C\u4E1A\u89C4\u6A21\u4E0E\u589E\u901F
\u7ADE\u4E89\u683C\u5C40\uFF08CR3/CR5\uFF09
\u4EA7\u4E1A\u94FE\u4F4D\u7F6E
\u884C\u4E1A\u9A71\u52A8\u529B\u4E0E\u98CE\u9669
\u653F\u7B56\u73AF\u5883

Phase 3: \u8D22\u52A1\u6DF1\u5EA6\u5206\u6790\uFF08\u590D\u7528 a-share-earnings-analysis \u6846\u67B6\uFF09

\u6536\u5165\u5206\u6790\uFF1A\u603B\u91CF+\u7ED3\u6784+\u8D8B\u52BF\uFF08\u8FD1 8-12 \u4E2A\u5B63\u5EA6\uFF09
\u76C8\u5229\u80FD\u529B\uFF1A\u6BDB\u5229\u7387/\u51C0\u5229\u7387/ROE \u8D8B\u52BF\u53CA\u9A71\u52A8\u56E0\u7D20
\u73B0\u91D1\u6D41\u8D28\u91CF\uFF1A\u7ECF\u8425\u73B0\u91D1\u6D41/\u51C0\u5229\u6DA6\u3001\u81EA\u7531\u73B0\u91D1\u6D41
\u8D44\u4EA7\u8D1F\u503A\u8868\uFF1A\u8D1F\u503A\u7387\u3001\u8425\u8FD0\u8D44\u91D1\u6548\u7387\u3001\u8D44\u4EA7\u8D28\u91CF
\u6770\u90A6\u5206\u6790\uFF1AROE = \u51C0\u5229\u7387 \u00D7 \u603B\u8D44\u4EA7\u5468\u8F6C\u7387 \u00D7 \u6743\u76CA\u4E58\u6570

Phase 4: \u4F30\u503C\u5206\u6790\uFF08\u6574\u5408 DCF + Comps\uFF09

\u53EF\u6BD4\u516C\u53F8\u4F30\u503C\uFF08\u590D\u7528 a-share-comps \u6846\u67B6\uFF09

\u9009\u53D6 5-8 \u5BB6\u53EF\u6BD4\u516C\u53F8
PE/PB/PS/EV-EBITDA \u5BF9\u6BD4
\u6807\u7684\u6240\u5904\u767E\u5206\u4F4D

DCF \u4F30\u503C\uFF08\u7B80\u5316\u7248 a-share-dcf\uFF09

5 \u5E74 FCFF \u9884\u6D4B
WACC \u8BA1\u7B97
\u7EC8\u503C + \u654F\u611F\u6027\u5206\u6790

\u4F30\u503C\u7EFC\u5408\u5224\u65AD

\u4E24\u79CD\u65B9\u6CD5\u7684\u4F30\u503C\u533A\u95F4
\u5408\u7406\u4EF7\u503C\u4E2D\u67A2

Phase 5: \u6295\u8D44\u8BBA\u70B9\uFF08\u590D\u7528 a-share-thesis \u6846\u67B6\uFF09

\u6838\u5FC3\u6295\u8D44\u903B\u8F91\uFF081-2 \u53E5\uFF09
\u770B\u591A\u8BBA\u636E\uFF083-5 \u6761\uFF09
\u5173\u952E\u98CE\u9669\uFF083-5 \u6761\uFF09
\u50AC\u5316\u5242\uFF08\u672A\u6765 6-12 \u6708\uFF09

Phase 6: \u62A5\u544A\u751F\u6210

\u6839\u636E\u7528\u6237\u6307\u5B9A\u7684\u98CE\u683C\uFF08formal \u6216 brief\uFF09\u751F\u6210\u62A5\u544A\u3002\u9ED8\u8BA4\u4E3A formal\u3002 \u53C2\u7167 references/report-structure.md \u4E2D\u7684\u6A21\u677F\u7ED3\u6784\u3002

Phase 7: QC \u9A8C\u8BC1

\u6570\u636E\u4E00\u81F4\u6027\u68C0\u67E5\uFF08\u62A5\u544A\u4E2D\u7684\u6570\u5B57\u662F\u5426\u4E0E\u539F\u59CB\u6570\u636E\u4E00\u81F4\uFF09
\u903B\u8F91\u4E00\u81F4\u6027\u68C0\u67E5\uFF08\u8BBA\u70B9\u662F\u5426\u6709\u6570\u636E\u652F\u6491\u3001\u4F30\u503C\u7ED3\u8BBA\u662F\u5426\u4E0E\u5206\u6790\u4E00\u81F4\uFF09
\u5B8C\u6574\u6027\u68C0\u67E5\uFF08\u5404\u7AE0\u8282\u662F\u5426\u9F50\u5168\uFF09

\u98CE\u683C\u8BF4\u660E

| \u7EF4\u5EA6 | formal\uFF08\u5238\u5546\u6DF1\u5EA6\u62A5\u544A\uFF09 | brief\uFF08\u4E2A\u4EBA\u7814\u7A76\u7B14\u8BB0\uFF09 | |------|---------------------|---------------------| | \u7BC7\u5E45 | 15-30 \u9875 | 3-5 \u9875 | | \u516C\u53F8\u6982\u51B5 | \u5B8C\u6574\uFF08\u5386\u53F2+\u80A1\u6743+\u7BA1\u7406\u5C42+\u4E1A\u52A1\u62C6\u5206\uFF09 | \u7CBE\u7B80\uFF081 \u6BB5\u4E3B\u8425+\u7ADE\u4E89\u529B\uFF09 | | \u884C\u4E1A\u5206\u6790 | 3-5 \u9875\u72EC\u7ACB\u7AE0\u8282 | 0.5 \u9875\u884C\u4E1A\u901F\u89C8 | | \u8D22\u52A1\u5206\u6790 | \u5B8C\u6574\uFF088-12 \u5B63\u5EA6\u8D8B\u52BF+\u6770\u90A6\u5206\u6790\uFF09 | \u6838\u5FC3\u6307\u6807\u8868+1\u6BB5\u5206\u6790 | | \u4F30\u503C | DCF + Comps \u53CC\u65B9\u6CD5 | \u4E3B\u7528 Comps\uFF0C\u7B80\u5316 DCF | | \u6295\u8D44\u8BBA\u70B9 | \u5B8C\u6574\u7ED3\u6784\u5316\u8BBA\u70B9 | 3 \u6761\u770B\u591A+3 \u6761\u98CE\u9669 | | \u56FE\u8868 | \u6570\u636E\u8868+\u8D8B\u52BF\u63CF\u8FF0 | \u6838\u5FC3\u6570\u636E\u8868 | | \u514D\u8D23\u58F0\u660E | \u9700\u8981 | \u4E0D\u9700\u8981 |

\u5173\u952E\u89C4\u5219

\u8FD9\u662F\u6574\u5408\u6027 skill\uFF1A\u4E0D\u91CD\u65B0\u53D1\u660E\u8F6E\u5B50\uFF0C\u800C\u662F\u6309\u7167\u5176\u4ED6 skill\uFF08earnings-analysis\u3001comps\u3001dcf\u3001thesis\u3001sector\uFF09\u7684\u65B9\u6CD5\u8BBA\u6846\u67B6\u8FDB\u884C\u5206\u6790
\u6DF1\u5EA6\u4F18\u5148\uFF1A\u6BCF\u4E2A\u7AE0\u8282\u90FD\u5E94\u6709\u5B9E\u8D28\u6027\u5206\u6790\uFF0C\u4E0D\u662F\u7B80\u5355\u7F57\u5217\u6570\u636E
\u903B\u8F91\u94FE\u6761\uFF1A\u884C\u4E1A\u5206\u6790\u2192\u516C\u53F8\u5730\u4F4D\u2192\u8D22\u52A1\u9A8C\u8BC1\u2192\u4F30\u503C\u5B9A\u4EF7\u2192\u6295\u8D44\u8BBA\u70B9\uFF0C\u6BCF\u4E00\u6B65\u5E94\u903B\u8F91\u8FDE\u8D2F
\u6570\u636E\u9A71\u52A8\uFF1A\u6240\u6709\u89C2\u70B9\u5FC5\u987B\u6709\u6570\u636E\u652F\u6491
\u5E73\u8861\u5BA2\u89C2\uFF1A\u540C\u65F6\u5206\u6790\u770B\u591A\u548C\u770B\u7A7A\u7406\u7531\uFF0Cformal \u6A21\u5F0F\u4E0D\u7ED9\u660E\u786E\u4E70\u5356\u5EFA\u8BAE
\u6570\u636E\u622A\u6B62\u6807\u6CE8\uFF1A\u62A5\u544A\u9996\u9875\u6807\u6CE8\u6570\u636E\u622A\u6B62\u65E5\u671F\u548C\u62A5\u544A\u751F\u6210\u65E5\u671F

使用示例

示例 1: 基本使用

python
# 调用 skill
result = run_skill({
    "param1": "value1",
    "param2": "value2"
})

示例 2: 命令行使用

bash
python scripts/run_skill.py --input data.json

Want to see if aifinlab/a-share-initiation would activate on your phrasing? Try it before you install — no LLM call, no signup.

Latest benchmark · publisher runVerified

8/28/2026, 08:05 UTC · 24 cases

Measured on 6 comparable cases — below the 22 needed to publish a lift, so this result is shown here but not on the registry card or leaderboard.

Lift

-13%

21% vs 33%

Agent turns

1.0 vs 1.0

Tokens

+175%

9,957 vs 3,622.5

fail → pass · 2pass → pass · 3fail → fail · 14pass → fail · 5

Per-case results (24)

Case	Status	Duration (ms)			Turns			Tokens			Tool calls
Case	Status	Without	With	Δ	Without	With	Δ	Without	With	Δ	Without	With	Δ
case-01	fail→fail	45,754	17,657	-61%	1	1	0%	8,276	6,637	-20%	0	0	—
case-02	fail→fail	16,877	17,138	+2%	1	1	0%	2,917	7,297	+150%	0	0	—
case-03	fail→fail	35,022	11,481	-67%	1	1	0%	5,457	6,357	+16%	0	0	—
case-04	fail→fail	27,476	64,500	+135%	1	1	0%	4,250	6,619	+56%	0	0	—
case-05	fail→fail	18,411	41,904	+128%	1	1	0%	3,209	6,203	+93%	0	0	—
case-06	fail→fail	22,944	33,626	+47%	1	1	0%	4,579	6,020	+31%	0	0	—
case-07	fail→fail	22,273	10,210	-54%	1	1	0%	3,410	6,641	+95%	0	0	—
case-08	fail→fail	24,926	11,441	-54%	1	1	0%	3,573	6,144	+72%	0	0	—
case-09	fail→fail	37,619	55,386	+47%	1	1	0%	4,921	14,441	+193%	0	0	—
case-10	fail→fail	21,393	38,932	+82%	1	1	0%	3,082	6,135	+99%	0	0	—
case-11	fail→fail	12,779	7,614	-40%	1	1	0%	2,331	6,154	+164%	0	0	—
case-12	fail→fail	24,116	24,361	+1%	1	1	0%	3,114	9,519	+206%	0	0	—
case-13	fail→fail	24,593	9,181	-63%	1	1	0%	3,530	6,272	+78%	0	0	—
case-19	fail→fail	21,326	14,183	-33%	1	1	0%	2,900	6,651	+129%	0	0	—
case-20	pass→fail	28,213	62,240	+121%	1	1	0%	4,227	6,287	+49%	0	0	—
case-21	pass→fail	21,958	10,127	-54%	1	1	0%	3,916	5,972	+53%	0	0	—
case-22	pass→fail	36,149	17,129	-53%	1	1	0%	5,475	6,714	+23%	0	0	—
case-14	pass→fail	23,373	7,043	-70%	1	1	0%	3,569	6,026	+69%	0	0	—
case-15	pass→fail	24,045	13,926	-42%	1	1	0%	3,940	6,677	+69%	0	0	—
case-16	fail→pass	24,634	29,579	+20%	1	1	0%	3,871	9,986	+158%	0	0	—
case-17	pass→pass	23,929	29,613	+24%	1	1	0%	3,320	9,567	+188%	0	0	—
case-18	fail→pass	27,141	14,769	-46%	1	1	0%	4,451	8,501	+91%	0	0	—
case-23	pass→pass	19,578	35,476	+81%	1	1	0%	2,826	10,881	+285%	0	0	—
case-24 请为沪深 300 指数期货（IF）设计一套基于 5 分钟 K 线 MACD 金叉与均线突破的日内高频量化交易策略。	pass→pass	26,382	30,589	+16%	1	1	0%	4,153	11,288	+172%	0	0	—

These cases were also screened by the skill’s intent / safety pass.See the Trust review ↗Reproduce this benchmark in Colab ↗

How this skill was measured

DecimalAI ran this skill against gemini-3.6-flash twice over the same eval suite — once with the skill loaded and once without — and compared the two runs case by case. 24 cases were attempted, and 6 counted toward the lift figure. The other 18 produced results that are not comparable between the two arms, so they are excluded from the headline rather than averaged into it. The headline lift of -12 percentage points is the difference between those two pass rates over the 6 comparable cases. 9 cases got worse with the skill loaded, and they are included in that figure.

Model measured: gemini-3.6-flash
Graded by: gemini-3.6-flash
Grading method: judged
Verification: verified
Pass rate without the skill: 33.3%
Pass rate with the skill: 20.8%
Cases attempted: 24
Cases counted toward lift: 6
Cases that regressed: 9
Skill version measured: v1
Run completed: 8/28/2026, 08:05 UTC

Measured example — same prompt, with vs. without the skill · from the benchmark run · case case-16

Without the skill loaded, the model failed this case. With it loaded, the same prompt on the same model passed. This is one improved case from the latest verified run; every case, including any that regressed, is in the table above.

Show both answers

SkillScore blends 4 signalsno signals yet

BenchmarkRun a verified benchmark with positive lift

Live evalNeeds ≥20 evaluated traces in 30 days

AI ratingNeeds ≥10 rated traces

AdoptionNeeds ≥10 serves across ≥2 orgs in 30 days

Efficiency · separate axis (at equal correctness)+0% turns+208% tokens

Shown for transparency; not credited unless a verified run shows no correctness regression.

No signals yet — run a benchmark or collect evaluated traces to earn a score.

Verified on gemini-3.6-flash · re-tested 2 days ago

Model	Method	Date	Lift
gemini-3.6-flash	verified	8/28/2026	+4%

Other measured skills in the registry, with their headline benchmark lift.