search-web

Decompose research goals into structured search queries with parallel subagent searches.

7|1|Updated Mar 30, 2024

Search Tags:#logs #search #web research #traceability #quality evaluation #multi-round #evaluator-optimizer

Ready to run

npx skills add https://github.com/Lionad-Morotar/simple-local-llm-server --skill search-web

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: search-web
Source: https://github.com/Lionad-Morotar/simple-local-llm-server/tree/main/local-link/skills/search-web
Command: npx skills add https://github.com/Lionad-Morotar/simple-local-llm-server --skill search-web

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) and references (resource) components.

What problem does it solve?

Web-scale research can be unreliable and time-consuming. This skill coordinates an evaluator-optimizer loop to structure searches, log results, and assess quality so you obtain verified findings with clear traceability.

Core Features & Use Cases

Structured planning: decompose goals into query clusters and run parallel searches with structured logs.
Quality-driven iteration: evaluation criteria (coverage, sources, consistency, specificity, recency) guide query optimization.
Reproducible deliverables: consolidated findings with sources and recommendations for research reports.

Quick Start

Define a research goal and launch the evaluator-optimizer loop to begin gathering and refining results.

Frequently Asked Questions about search-web

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

FAQPage Schema

How do I structure web research queries for traceable and reproducible results?▼

You can structure web research by decomposing goals into query clusters and running parallel searches. This evaluator-optimizer loop logs results and assesses quality to ensure findings are traceable and reproducible.

What is an evaluator-optimizer loop for web research?▼

An evaluator-optimizer loop for web research is a mechanism that iteratively improves search findings. It applies criterion-based evaluation covering coverage, sources, consistency, specificity, and recency to guide query optimization.

How do I evaluate the quality of multi-round search findings?▼

To evaluate multi-round search findings, apply structured criteria such as coverage, source consistency, specificity, and recency. These evaluation metrics guide iterative query optimization until quality requirements are satisfied.

When do I need structured logging for search traceability?▼

Structured logging for search traceability is needed when consolidating research deliverables. It ensures sources are tracked and results are reproducible, providing clear handoff criteria to decide when to complete the research.

Can I use parallel subagent searches to decompose a research goal?▼

Yes, you can use parallel subagent searches to decompose a research goal. This approach breaks down the overarching objective into actionable components and generates structured search queries to gather findings simultaneously.