web-search-plus_skill

This skill automatically routes queries to the best among Serper, Tavily, Exa, You.com, and SearXNG for fast, accurate web results.

Python

5

GitHub Stars

2

Bundled Files

2 months ago

Catalog Refreshed

4 months ago

First Indexed

Readme & install

Copy the install command, review bundled files from the catalogue, and read any extended description pulled from the listing source.

Installation

Preview and clipboard use veilstrat where the catalogue uses aiagentskills.

npx veilstrat add skill robbyczgw-cla/web-search-plus --skill web-search-plus

README.md22.5 KB
SKILL.md8.2 KB

Overview

This skill unifies multiple search providers and automatically selects the best one for each query using multi-signal analysis and confidence scoring. It supports Serper (Google), Tavily (research), Exa (neural similarity), You.com (RAG/real-time), and SearXNG (privacy/self-hosted) behind a single interface. Auto-routing reduces provider choice friction while allowing explicit overrides and fallback handling.

How this skill works

The skill analyzes each query for intent, keywords, URL patterns, and real-time signals, then ranks providers by suitability and confidence. It dispatches the query to the selected provider, returns standardized results with a routing block (provider, confidence, fallback flags), and can automatically retry alternative providers on errors. Developers can enable explain-mode to see routing rationale or force a provider via CLI flags or API parameters.

When to use it

General web queries without choosing a provider manually
Shopping, local, or news lookups (auto-routes to Serper)
Academic or deep research questions (auto-routes to Tavily)
Finding similar pages, startups, or semantic matches (auto-routes to Exa)
Privacy-first searches or cost-free/self-hosted needs (use SearXNG)

Best practices

Start with one API key and add providers progressively for coverage
Use auto-routing for most queries; override with -p when you need a specific engine
Enable --explain-routing to debug ambiguous routing decisions
Configure a sensible fallback_provider and set a confidence_threshold for stricter routing
Self-host SearXNG for private, unlimited queries and set it as fallback if budget is a concern

Example use cases

Price comparison: ask “iPhone 16 price” and get Serper-powered shopping results
Research deep topics: “climate change research 2024” routes to Tavily for scholarly context
Similarity search: “companies like Stripe” uses Exa for neural matches
Breaking news or real-time updates: “latest AI regulation news” routes to You.com
Private searches or cost-free fallback: route sensitive queries to a self-hosted SearXNG instance

FAQ

No. You only need keys for the providers you plan to use. The skill works with a single key and can add more later.

What if auto-routing chooses the wrong provider?

Use the -p provider override or run the tool with --explain-routing to see why it chose a provider and adjust query or config.