skills.rest
CYP0630
CYP0630COMMUNITY
Explore all skills by this author →
WebsiteGitHub

trading

Decide BUY, SELL, or HOLD for a symbol and date using DuckDB data.

Updated May 3, 2026
Search Tags:#automation#mcp#trading#duckdb#upsert#equities
One-click install
npx skills add https://github.com/CYP0630/agentic_benchmark_single_agent --skill trading-cyp0630
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: trading
Source: https://github.com/CYP0630/agentic_benchmark_single_agent/tree/main/skills/trading
Command: npx skills add https://github.com/CYP0630/agentic_benchmark_single_agent --skill trading-cyp0630

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires duckdb, pandas, pandas_ta, numpy, fastmcp, pydantic, and includes scripts (resource) components.

What problem does it solve?

Makes a single-day trading decision for one symbol on a given target_date by querying an offline DuckDB via MCP tools, and upserting the result into an action-list JSON file. The skill is driven externally: each invocation handles exactly one (symbol, target_date) pair and uses MCP to access data, enabling straightforward backtests and live trading.

Core Features & Use Cases

  • Single-day decision per symbol and date and upsert into an action-list JSON file.
  • Data access via the trading_mcp server to read prices, news, filings, and indicators from an offline DuckDB.
  • Supports backtest replay and live trading by iterating dates or using the latest data.
  • Example phrases: trade AAPL on 2025-03-05, decide TSLA today, run trading for MSFT 2025-04-10.

Quick Start

Trade a stock on a specific date by asking the agent to run the trading action for that symbol and date.

Frequently Asked Questions about trading

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I make a single-day stock trading decision using offline data?▼

To make a single-day stock trading decision, the skill queries offline DuckDB data via MCP to read prices, news, filings, and indicators, then outputs a BUY, SELL, or HOLD action for a specific symbol and date.

Can I backtest equity trading strategies with DuckDB and MCP?▼

Yes, you can backtest equity trading strategies by iterating through historical dates. Each invocation processes one symbol and date pair, upserting the trading decision into an action-list JSON file for replay.

How does the upsert mechanism work for trading action lists?▼

The upsert mechanism updates or inserts a trading decision entry into an action-list JSON file. It records the target date and price for a specific symbol, ensuring the action list remains current for both backtesting and live trading.

Do I need a local DuckDB server to run technical indicator analysis for trading?▼

Yes, you need the trading_mcp server to access a local DuckDB database. This setup allows the skill to read historical prices, news, filings, and technical indicators offline before deciding a trading action.

What technical indicators are available for equity trading decisions?▼

The skill uses pandas_ta to calculate technical indicators from offline DuckDB data. These indicators, alongside prices, news, and filings, are accessed via MCP to inform the BUY, SELL, or HOLD decision for the target date.

Related Skills

Explore More Finance & Accounting Skills→
xueqingpengxueqingpeng
11

report_generation

Generate a weekly equity research report from offline DuckDB data.

Community
Advanced
sarkcesscrewpaysarkcesscrewpay

stock-analysis

Analyze stock and crypto tickers to generate buy/hold/sell signals with Yahoo Finance data.

Community
Intermediate
jayjpatel9717jayjpatel9717

technical-analysis

Automate technical analysis on stock market data using pandas and numpy.

Community
Intermediate
agiprolabsagiprolabs
266

vectorbt

Backtest trading strategies with vectorized vectorbt simulations and grid search optimization.

Community
Advanced