global‑stock‑data – Market‑data skill for AI coding assistants

What it is

• A single Skill file (Markdown with embedded Python) that lets AI coding assistants (Claude Code, OpenAI Codex, OpenClaw, etc.) fetch a huge variety of US‑focused market data on‑the‑fly.
• The repository contains only the code that talks to public data sources; the actual data is pulled by the user at request time.

Why it matters

• AI assistants often need up‑to‑date financial information to answer coding‑related questions (e.g., “show AAPL’s 0‑day‑to‑expiry options flow”).
• Instead of each assistant developer re‑implementing dozens of scrapers, this skill provides a ready‑made, zero‑auth, requests‑only interface covering 16 logical layers (quotes, K‑lines, fundamentals, options Greeks, short‑interest, SEC filings, macro rates, FX, etc.) and 50+ endpoints.

Key features (as described in the README)

Layer	Main data & sources	Highlights
Market Data	Sina, Tencent, Eastmoney (real‑time US/HK quotes)	25‑78 fields per symbol
K‑line / chart	Sina (back to 1984) + Yahoo (global)	Daily, weekly, monthly, minute bars; works for US, JP, TW, KR, IN, EU markets
Technical indicators	Pure‑Python calculations	MA, EMA, MACD, RSI, KDJ, Bollinger
Fundamentals	Eastmoney, Yahoo, SEC XBRL	Income‑statement, balance‑sheet, cash‑flow, 503 GAAP metrics
Fund flow	Eastmoney push2his	Daily flow by market‑cap tier
Options	Yahoo crumb (fallback)	Full chain for US symbols
Official‑source options	CBOE (cdn.cboe.com)	Full chain + implied vol + Greeks + 0‑day‑to‑expiry (0DTE) flow detection
Short volume	FINRA Reg SHO	Market‑wide daily short volume for ~12 k symbols
Short interest	FINRA (twice‑monthly)	Open short positions, days‑to‑cover, rankings for ~22 k symbols
SEC filings	EDGAR daily index + XBRL	Same‑day Form 4, 8‑K, 13F; full‑text search back to 2001
Insider trades / 13F	SEC Form 4 & 13F parsers	Line‑by‑line insider trade details, quarterly institutional holdings
Macro / calendar	US Treasury, CFTC, Nasdaq earnings calendar	Yield curve (1 M‑30 Y), COT positions, earnings dates
Global rates & FX (V2.1)	BIS, ECB, NY Fed, Fed Board, Treasury	Policy rates for 38 central banks, EUR/USD, SOFR/EFFR, euro‑area yield curve
Taiwan market (V2.2)	TWSE & TPEx open APIs	Latest quotes, P/E, dividend yield, monthly revenue, margin data
Energy (V2.2)	US EIA	Weekly crude/gas inventories, daily spot prices for WTI, Brent, Henry Hub
Tools	Various	Symbol search, market‑wide ticker list, news, ticker↔CIK mapping

Compliance tiers

• The README grades every data source from S (free, commercial‑friendly) to D (personal‑use only) and flags those that are B (need permission for commercial reuse) or C/D (prohibited for automated scraping). This helps developers decide which endpoints are safe for production use.

How to get started

1. Create a ~/.claude/skills/global-stock-data directory.
2. Download the single SKILL.md file from the repo.
3. Install the only runtime dependency: pip install requests (Windows also needs tzdata).
4. (Optional) Fill in SEC_CONTACT with your name/email for SEC‑required User‑Agent.
5. Invoke the skill from an AI assistant, e.g., “Claude, show AAPL’s financials”.

Typical use‑case prompts (from the README)

• “Show NVDA’s 0DTE options flow – unusual activity.”
• “TSLA’s short‑volume ratio trend this week.”
• “Current policy rates of the major central banks.”
• “WTI price chart for the past year.”
• “Berkshire’s top 10 holdings in its latest 13F.”

Who it’s for

• Developers building AI‑assistant‑driven finance tools.
• Quant researchers who want a quick, code‑only way to pull public market data without dealing with dozens of APIs.
• Anyone needing a zero‑auth, Python‑only wrapper around publicly available financial feeds.

Limitations

• The repository supplies only the code; you must respect each source’s terms (especially tiers C/D). Commercial projects should stick to tier S sources.
• Some data (e.g., Yahoo, Nasdaq) is technically disallowed for automated scraping; the skill includes the code but warns users.

Bottom line global-stock-data is a well‑documented, open‑source “skill” that aggregates a broad set of public US and global market data sources into a single, AI‑assistant‑friendly Python interface, with clear compliance guidance.