Is this skill a callable function for searching stocks?

No, it is a guidance protocol. It instructs the agent on how to use search tools like MCP Exa Web Search to perform searches, extract data, and perform cross-verification manually.

What happens if the data from two sources does not match?

If the variance is within 5%, the average is used. If it is between 5-10%, the most recent source is prioritized with a warning. If variance exceeds 10%, the verification returns a FAIL status.

Does it support real-time price data?

Yes, it guides the agent to fetch the most recent data available via web search while enforcing a 7-day freshness limit to ensure the price information is not outdated.

Why is the 'Original Text Citation' mandatory?

Requiring the exact sentence from the source ensures the AI has not hallucinated the number and provides a verifiable audit trail for the human user to check the context of the data.

Which financial markets does this skill support?

The protocol includes specific search patterns and data sources for the South Korean (KOSPI/KOSDAQ) and United States (NYSE/NASDAQ) markets, as well as global ETFs.

Stock Data Verifier

Name: Stock Data Verifier
Author: orientpine

byorientpine

•

ウェブスクレイピングとデータ収集

Implements a rigorous cross-verification protocol for stock and ETF data to prevent AI hallucinations through multi-source validation.

The Stock Data Verifier is a specialized protocol designed to ensure the highest level of accuracy when AI agents retrieve financial data. It establishes a 'Zero Tolerance' policy for hallucinations by requiring agents to manually invoke web search tools, collect data from at least two independent high-authority sources (such as KRX, Yahoo Finance, or official fund providers), and verify that values fall within a strict 5% margin of error. By mandating the inclusion of original text citations for every numerical value, this skill transforms Claude into a reliable financial analyst capable of providing audited, up-to-date market information.

主な機能

01Standardized JSON output schemas for verified stock and ETF metrics

02Automated variance calculation and ±5% error tolerance enforcement

03Multi-source cross-validation protocol for KR and US markets

04Specific search query patterns for high-authority financial domains

05Mandatory 'Original Text' citation requirement for all numerical data

069 GitHub stars

ユースケース

01Verifying real-time ETF expense ratios and NAV data from official fund providers

02Generating high-reliability automated stock market analysis reports

03Auditing financial multiples like PER and PBR across multiple global data sources

主な機能

01Standardized JSON output schemas for verified stock and ETF metrics

02Automated variance calculation and ±5% error tolerance enforcement

03Multi-source cross-validation protocol for KR and US markets

04Specific search query patterns for high-authority financial domains

05Mandatory 'Original Text' citation requirement for all numerical data

069 GitHub stars

ユースケース

01Verifying real-time ETF expense ratios and NAV data from official fund providers

02Generating high-reliability automated stock market analysis reports

03Auditing financial multiples like PER and PBR across multiple global data sources