analyst-common

Enforce web-search calls, verbatim quotes, and cross-source verification for AI analysts.

Updated Jan 22, 2026
One-click install
npx skills add https://github.com/ByungJu-Lim/obsidian-- --skill analyst-common-byungju-lim
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: analyst-common
Source: https://github.com/ByungJu-Lim/obsidian--/tree/main/0-Projects/honeypot-main/honeypot-main/plugins/investments-portfolio/skills/analyst-common
Command: npx skills add https://github.com/ByungJu-Lim/obsidian-- --skill analyst-common-byungju-lim

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

AI analysts often hallucinate or misinterpret data when synthesizing information from the web. This skill enforces direct web-search usage, requires verbatim quotes of numeric results, and mandates cross-source verification to ensure accuracy and traceability.

Core Features & Use Cases

  • Direct web-search tool invocation is mandatory, ensuring the agent relies on live results.
  • Verbatim quotation of numeric data with source URLs and dates for auditable provenance.
  • Cross-source validation across at least three independent sources with a ±1% agreement threshold to confirm data consistency.
  • Applicable to index-fetcher, rate-analyst, sector-analyst, risk-analyst, leadership-analyst, and macro-critic workflows requiring reliable web data.

Quick Start

Instruct the agent to call mcp_websearch_web_search_exa for a query and record the original_text, numeric value, and three reliable sources with URLs and dates.

Frequently Asked Questions about analyst-common

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

FAQPage Schema
How do I prevent AI hallucinations during web research and data analysis?▼

To prevent AI hallucinations during web research, you can enforce direct web-search calls, require verbatim quotations of numeric data, and mandate cross-source verification across at least three independent sources with a ±1% numeric agreement threshold.

How do I cross-validate numeric data from web search results?▼

Cross-validate numeric data by capturing the original_text verbatim, acquiring at least three independent sources with recorded URLs and dates, and checking that the numeric values agree within a strict ±1% threshold.

What is the best way to ensure source citation traceability for AI analyst workflows?▼

The best way to ensure source citation traceability is to enforce mandatory web-search tool invocation and record verbatim original_text alongside the exact source URLs and retrieval dates for every numeric data point cited.

Can I use cross-source verification for sector and risk analysis tasks?▼

Yes, cross-source verification can be applied across sector-analyst, risk-analyst, rate-analyst, and macro-critic workflows to ensure reliable web data and prevent misinterpretation during information synthesis.

Why does my AI analyst hallucinate when synthesizing market data from the web?▼

AI analysts hallucinate when synthesizing market data because they lack enforced direct web-search usage and verbatim quotation requirements, leading to unverified data interpretation without traceable source provenance.

What are the limitations of relying on a single web source for numeric data extraction?▼

Relying on a single web source for numeric data extraction lacks cross-source validation, meaning you cannot confirm data consistency within a ±1% agreement threshold or ensure the accuracy required to prevent AI hallucinations.