Prompt file imported from GoatWang/oysterun-ccstock-workenv (.codex/prompts/screen_stocks.md). Copyright stays with the author.

篩選股票

Name: screen_stocks
Author: GoatWang

Screen stocks based on financial criteria using natural language.

Usage

/screen_stocks <criteria>

Examples

/screen_stocks P/E < 15 且 ROE > 15%
/screen_stocks 台股中殖利率 > 5% 的股票
/screen_stocks US stocks with market cap > 10B and D/E < 0.5
/screen_stocks 找出三年低點的中國A股

Instructions

When this command is invoked:

Parse natural language criteria into structured filters:
- Identify market (US/TW/CN) from context
- Extract metric conditions (P/E, ROE, P/B, D/E, yield, market cap, etc.)
- Detect special patterns like "N-year low", "near 52-week low", etc.
Run the screener:

cd $(pwd)/tool_scripts/financial_calc && uv run python screener.py --market <market> --criteria '<json_criteria>'

Where json_criteria is a JSON string like:

{"market": "US", "filters": [{"metric": "pe_ratio", "op": "<", "value": 15}, {"metric": "roe", "op": ">", "value": 0.15}]}

Format output in 繁體中文:

🔍 篩選條件：<human-readable criteria>
市場：<market>
符合條件：<count> 檔

| # | 代號 | 名稱 | 現價 | P/E | ROE | <relevant metrics> |
|---|------|------|------|-----|-----|--------------------|
| 1 | XXXX | XXX  | XX.X | XX  | XX% | ...                |
...

📊 共 <count> 檔符合條件

If results exceed 20, show top 20 and mention total count.
Report the formatted result to the user.

{ "prompt": "# 篩選股票\n\nScreen stocks based on financial criteria using natural language.\n\n## Usage\n\n```\n/screen_stocks <criteria>\n```\n\n## Examples\n\n```\n/screen_stocks P/E < 15 且 ROE > 15%\n/screen_stocks 台股中殖利率 > 5% 的股票\n/screen_stocks US stocks with market cap > 10B and D/E < 0.5\n/screen_stocks 找出三年低點的中國A股\n```\n\n## Instructions\n\nWhen this command is invoked:\n\n1. **Parse natural language criteria** into structured filters:\n - Identify market (US/TW/CN) from context\n - Extract metric conditions (P/E, ROE, P/B, D/E, yield, market cap, etc.)\n - Detect special patterns like \"N-year low\", \"near 52-week low\", etc.\n\n2. **Run the screener:**\n\n```bash\ncd $(pwd)/tool_scripts/financial_calc && uv run python screener.py --market <market> --criteria '<json_criteria>'\n```\n\nWhere `json_criteria` is a JSON string like:\n```json\n{\"market\": \"US\", \"filters\": [{\"metric\": \"pe_ratio\", \"op\": \"<\", \"value\": 15}, {\"metric\": \"roe\", \"op\": \">\", \"value\": 0.15}]}\n```\n\n3. **Format output in 繁體中文:**\n\n```\n🔍 篩選條件：<human-readable criteria>\n市場：<market>\n符合條件：<count> 檔\n\n| # | 代號 | 名稱 | 現價 | P/E | ROE | <relevant metrics> |\n|---|------|------|------|-----|-----|--------------------|\n| 1 | XXXX | XXX | XX.X | XX | XX% | ... |\n...\n\n📊 共 <count> 檔符合條件\n```\n\n4. If results exceed 20, show top 20 and mention total count.\n\n5. Report the formatted result to the user.", "variables": [], "notes": "Screen stocks based on financial criteria using natural language.", "metadata": { "source": { "provider": "github-codex-prompts", "repository": "https://github.com/GoatWang/oysterun-ccstock-workenv", "path": ".codex/prompts/screen_stocks.md", "ref": "251a8b15e44d108116b78783c5941faf94dcc861", "url": "https://github.com/GoatWang/oysterun-ccstock-workenv/blob/251a8b15e44d108116b78783c5941faf94dcc861/.codex/prompts/screen_stocks.md", "key": "GoatWang/oysterun-ccstock-workenv/.codex/prompts/screen_stocks.md" } } }

screen_stocks

About

篩選股票

Usage

Examples

Instructions

Use it

Manifest

Reviews