finviz-screener

FinViz Screener

Overview

Translate natural-language stock screening requests into FinViz screener filter codes, build the URL, and open it in Chrome. No API key required for public screener; FINVIZ Elite is auto-detected from $FINVIZ_API_KEY for enhanced functionality.

Key Features:

• Natural language → filter code mapping (Japanese + English)
• URL construction with view type and sort order selection
• Elite/Public auto-detection (environment variable or explicit flag)
• Chrome-first browser opening with OS-appropriate fallbacks
• Strict filter validation to prevent URL injection

When to Use This Skill

Explicit Triggers:

• "高配当で成長している小型株を探したい"
• "Find oversold large caps near 52-week lows"
• "テクノロジーセクターの割安株をスクリーニングしたい"
• "Screen for stocks with insider buying"
• "FinVizでブレイクアウト候補を表示して"
• "Show me high-growth small caps on FinViz"
• "配当利回り5%以上でROE15%以上の銘柄を探して"

Implicit Triggers:

• User describes stock screening criteria using fundamental or technical terms
• User mentions FinViz screener or stock filtering
• User asks to find stocks matching specific financial characteristics

When NOT to Use:

• Deep fundamental analysis of a specific stock (use us-stock-analysis)
• Portfolio review with holdings (use portfolio-manager)
• Chart pattern analysis on images (use technical-analyst)
• Earnings-based screening (use earnings-trade-analyzer or pead-screener)

Workflow

Step 1: Load Filter Reference

Read the filter knowledge base:

cat references/finviz_screener_filters.md

Step 2: Interpret User Request

Map the user's natural-language request to FinViz filter codes. Use the Common Concept Mapping table below for quick translation, and reference the full filter list for precise code selection.

Note: For range criteria (e.g., "dividend 3-8%", "P/E between 10 and 20"), use the {from}to{to} range syntax as a single filter token (e.g., fa_div_3to8, fa_pe_10to20) instead of combining separate _o and _u filters.

Common Concept Mapping:

Step 3: Present Filter Selection

Before executing, present the selected filters in a table for user confirmation:

| Type | Value | Meaning |
|---|---|---|
| Theme | artificialintelligence | Artificial Intelligence |
| Sub-theme | aicloud | AI - Cloud & Infrastructure |
| Filter | cap_small | Small Cap ($300M–$2B) |
| Filter | fa_div_o3 | Dividend Yield > 3% |

Ask the user to confirm or adjust before proceeding.

Step 4: Execute Script

Run the screener script to build the URL and open Chrome:

Script arguments:

• --filters (optional): Comma-separated filter codes. Note: theme_* and subtheme_* tokens are not allowed here — use --themes / --subthemes instead.
• --themes (optional): Comma-separated theme slugs (e.g., artificialintelligence,cybersecurity). Accepts bare slugs or theme_-prefixed values.
• --subthemes (optional): Comma-separated sub-theme slugs (e.g., aicloud,aicompute). Accepts bare slugs or subtheme_-prefixed values.
• --elite: Force Elite mode (auto-detected from $FINVIZ_API_KEY if not set)
• --view: View type — overview, valuation, financial, technical, ownership, performance, custom
• --order: Sort order (e.g., -marketcap, dividendyield, -change)
• --url-only: Print URL without opening browser

At least one of --filters, --themes, or --subthemes must be provided.

Step 5: Report Results

After opening the screener, report:

1. The constructed URL
2. Elite or Public mode used
3. Summary of applied filters
4. Suggested next steps (e.g., "Sort by dividend yield", "Switch to Financial view for detailed ratios")

Usage Recipes

Real-world screening patterns distilled from repeated use. Each recipe includes a starter filter set, recommended view, and tips for iterative refinement.

Recipe 1: High-Dividend Growth Stocks (Kanchi-Style)

Goal: High yield + dividend growth + earnings growth, excluding yield traps.

Iterative refinement: Start broad with fa_div_o3 → review results → add fa_div_3to8 to cap yield → add fa_payoutratio_u60 to exclude traps → switch to financial view for payout and growth columns.

Recipe 2: Minervini Trend Template + VCP

Goal: Stocks in a Stage 2 uptrend with volatility contraction (VCP setup).

VCP tightening filters (add to narrow): ta_volatility_wo3,ta_highlow20d_b0to5h,sh_relvol_u1 — low weekly volatility, near 20-day high, below-average relative volume (contraction signal).

Recipe 3: Unfairly Sold-Off Growth Stocks

Goal: Fundamentally strong companies with recent sharp declines — potential mean reversion candidates.

After review: Switch to valuation view to check P/E and P/S for entry attractiveness.

Recipe 4: Turnaround Stocks

Goal: Companies with previously declining earnings now showing recovery — bottom-fishing with fundamental confirmation.

Recipe 5: Momentum Trade Candidates

Goal: Short-term momentum leaders near 52W highs with increasing volume.

Recipe 6: Theme Screening (AI + Sub-theme Drill-Down)

Goal: Find mid-cap+ AI stocks focused on cloud infrastructure and compute acceleration.

Multi-theme example: --themes "artificialintelligence,cybersecurity" selects stocks tagged with either theme (OR logic via | grouping).

Tips: Iterative Refinement Pattern

Screening works best as a dialogue, not a one-shot query:

1. Start broad — use 3-4 core filters to get an initial result set
2. Review count — if too many results (>100), add tightening filters; if too few (<5), relax constraints
3. Switch views — start with overview for a quick scan, then switch to financial or valuation for deeper inspection
4. Layer in technicals — after confirming fundamental quality, add ta_ filters to time entries
5. Save and iterate — bookmark the URL, then adjust one filter at a time to understand its impact

Resources

• references/finviz_screener_filters.md — Complete filter code reference with natural language keywords (includes industry code examples; full 142-code list is in the Industry Codes section)
• scripts/open_finviz_screener.py — URL builder and Chrome opener