multi-source-inquiry

Cross-verify claims across multiple source categories with evidence labels.

8|2|Updated Oct 29, 2025
One-click install
npx skills add https://github.com/cuipengfei/prompts --skill multi-source-inquiry
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: multi-source-inquiry
Source: https://github.com/cuipengfei/prompts/tree/main/plugins/deep-research/skills/multi-source-inquiry
Command: npx skills add https://github.com/cuipengfei/prompts --skill multi-source-inquiry

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill helps you run research tasks that need联网与多来源核查 so you can produce reliable, well-structured conclusions instead of single-source guesses.

Core Features & Use Cases

  • Multi-source fact-checking: evaluates source volume, selects the smallest sufficient workflow, and cross-verifies claims across different source types.
  • MECE-based large-scope decomposition: splits big topics into 3-6 boundary-clear sub-tasks that can be delegated and then merged without redundant full searches.
  • Evidence-first outputs with conflict handling: traces claims back to original sources, performs反方查询 to downgrade or mark [conflicting], and labels certainty as [verified], [likely], [single-source], or [unknown].

Quick Start

Use the multi-source-inquiry skill to fact-check this claim: "X causes Y" by comparing at least two high-credibility source categories and outputting a verified/likely/conflicting evidence summary with URLs.

Frequently Asked Questions about multi-source-inquiry

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

FAQPage Schema
How do I fact-check claims using multi-source web research?

Multi-source fact-checking evaluates source volume and cross-verifies claims across at least two high-credibility source categories. It applies SIFT-style sourcing to produce evidence-backed summaries with URLs.

What is MECE decomposition for complex research workflows?

MECE decomposition splits large research topics into 3-6 boundary-clear sub-tasks. These sub-tasks can be delegated to parallel sub-agents and merged without redundant full searches.

How does conflict resolution work for source discrepancies?

Conflict resolution traces claims back to original sources and performs anti-claim querying. Conflicting evidence is downgraded or marked as `[conflicting]` to ensure accurate synthesis.

How to label evidence certainty during fact checking?

Evidence certainty is labeled as `[verified]`, `[likely]`, `[single-source]`, or `[unknown]`. This evidence-first output format ensures conclusions are traced back to original sources with clear confidence levels.

Can I use parallel sub-agents for large-scope research decomposition?

Yes, MECE-based large-scope decomposition supports optional parallel sub-agents. It splits big topics into boundary-clear sub-tasks that can be delegated and then merged without redundant full searches.

When do I need SIFT-style sourcing for evidence synthesis?

SIFT-style sourcing is required for structured multi-source research and claim verification. It ensures cross-validation across different source types and supports evidence-labeled outputs with conflict detection.