Finter Skills for Claude Code

Skills for quant research with Finter platform in Claude Code.

Quick Start

# Clone and install
git clone git@github.com:Quantit-Github/finter-skills.git
cd finter-skills
./install.sh

# Start working
cd workspace
claude

What install.sh Does

finter-skills/
├── finter-data/          # Skills (source)
├── finter-alpha/
├── finter-portfolio/
├── ...
├── install.sh
├── CLAUDE.md
└── workspace/            # <- Your working directory
    ├── .claude/
    │   └── skills/       # Symlinks to ../finter-*
    ├── CLAUDE.md
    ├── pyproject.toml
    └── .venv/

Usage Examples

> What data is available for Korean stocks?

> Create a momentum alpha for us_stock

> Backtest a BTC strategy with crypto_test universe

> Implement a value factor for kr_stock

Universes

Universe	Assets	Best For
kr_stock	~2,500	Korean stocks (richest data)
us_stock	~8,000	US stocks
us_etf	~6,700	US ETFs
vn_stock	~1,000	Vietnam stocks
id_stock	~1,000	Indonesia stocks
crypto_test	~378	Crypto (2024 only, 10min candles)

Skills

Skill	Description
finter-data	Data loading, Symbol search, preprocessing
finter-alpha	BaseAlpha strategy, backtesting
finter-portfolio	Portfolio optimization
finter-insight	Research DB search (RAG)
finter-operations	Error recovery workflow

Requirements

• macOS / Linux
• Python 3.10+
• Claude Code CLI (claude)