Web Search Plus
Run unified web search and URL extraction across many providers with auto-routing so agents pick the right API without manual provider shopping.
Overview
Web Search Plus is a journey-wide agent skill that routes web search and URL extraction across 11 providers—usable whenever a solo builder needs live web evidence before committing to a plan.
Install
npx skills add https://github.com/robbyczgw-cla/web-search-plus --skill web-search-plusWhat is this skill?
- Intelligent auto-routing across 11 search providers plus companion URL extraction
- Providers include Serper, Brave, Tavily, Querit, Linkup, Exa, Firecrawl, Perplexity, You.com, SearXNG, and SerpBase
- Search needs one provider key or self-hosted SearXNG; extraction uses Firecrawl, Linkup, Tavily, Exa, or You.com
- Supports multilingual search, shopping-oriented queries, RAG-style fetch, privacy/self-hosted options, and free-tier pat
- Requires python3 and bash; API keys are optional per provider with documented env vars
- 11 search providers with intelligent auto-routing
- Companion extraction flow across Firecrawl, Linkup, Tavily, Exa, or You.com
Adoption & trust: 1.2k installs on skills.sh; 21 GitHub stars; 1/3 security scanners passed (skills.sh audits).
What problem does it solve?
You have five different search API keys and still waste agent turns deciding which provider fits broad web vs semantic research vs page extraction.
Who is it for?
Agent builders who need dependable live web data across idea validation, implementation research, and competitive monitoring with one skill install.
Skip if: Workflows that are entirely offline, or teams forbidden from sending queries to third-party search APIs without legal review.
When should I use this skill?
User needs web search, multilingual SERP results, semantic research, shopping queries, or URL content extraction with provider auto-routing.
What do I get? / Deliverables
Your agent runs a single search-plus-extract workflow with documented env keys, bash/python tooling, and provider auto-routing instead of bespoke scripts per vendor.
- Routed search results from the best-fit configured provider
- Extracted URL text suitable for agent context or RAG
- Documented env var setup for chosen vendors
Recommended Skills
Journey fit
Useful at every journey phase - explore requirements and options before committing to a direction.
Where it fits
Compare competitor positioning and pricing pages before choosing a niche.
Pull official docs snippets to confirm an integration is feasible for an MVP.
Verify current API behavior and changelog entries while wiring a third-party SDK.
Research SERP snippets and FAQ patterns for a launch landing page.
Search forums and recent articles when users report breakage after an upstream change.
How it compares
Unified multi-provider search skill—not a single-vendor MCP-only wrapper or manual curl recipes.
Common Questions / FAQ
Who is web-search-plus for?
Solo and indie builders running coding agents that must search the open web and extract page content without maintaining separate Serper, Tavily, and Firecrawl playbooks.
When should I use web-search-plus?
In Idea for market and competitor research; in Validate for pricing and positioning facts; in Build when verifying APIs and docs; in Ship/Launch for release notes and policy checks; in Grow for trend and support forum scans—whenever live URLs beat model memory.
Is web-search-plus safe to install?
Check the Security Audits panel on this Prism page; the skill uses network access and optional third-party API keys, so review data handling for each provider you enable.
SKILL.md
READMESKILL.md - Web Search Plus
# Web Search Plus **Stop choosing search providers. Let the skill do it for you.** This skill now connects you to **11 search providers** and adds a companion extraction flow for pulling content from URLs. Broad web query? → Brave or Serper. Research question? → Tavily or Exa. Need citations and grounding? → Linkup. Want scrape-ready content? → Firecrawl. Prefer privacy? → SearXNG. Need low-cost Google SERP with prepaid credits? → SerpBase (explicit/fallback-only). --- ## ✨ What Makes This Different? - **Just search** — no need to think about which provider to use - **Smart routing** — query analysis picks the best provider automatically - **11 providers, 1 interface** — general web, research, semantic discovery, direct answers, privacy-first, prepaid-credits, and extraction-capable providers together - **URL extraction included** — pull markdown/HTML content with fallback across five providers - **Works with just 1 credential** — start with any single provider, add more later - **Free/self-hosted options available** — SearXNG can run at $0 API cost --- ## 🚀 Quick Start ```bash # Interactive setup (recommended for first run) python3 scripts/setup.py # Or manually cp .env.example .env python3 scripts/search.py -q "latest OpenClaw release" python3 scripts/extract.py --url https://example.com ``` The wizard explains providers, collects keys, and sets defaults. --- ## 🔑 Providers ### Search providers - **Serper** — shopping, prices, local, and general Google-style results; fast broad fallback - **Brave** — independent web index and generic current-web queries; strong non-Google complement - **Tavily** — research, explanations, and synthesis; strong research routing - **Querit** — multilingual and international updates; good for cross-language recency - **Linkup** — source-grounded/citation-heavy search; evidence-first queries - **Exa** — semantic discovery, similar sites, and deep research; supports `deep` + `deep-reasoning` - **Firecrawl** — search with scrape-ready metadata; also strong extraction provider - **Perplexity** — direct answers with citations; via `PERPLEXITY_API_KEY` or `KILOCODE_API_KEY` - **You.com** — current-web / RAG-friendly snippets; also supports extraction - **SearXNG** — private/self-hosted search; no API key, just instance URL - **SerpBase** — low-cost Google SERP with prepaid credits; explicit/fallback-only (opt-in via `--provider serpbase` or add to `provider_priority` in `config.json`) ### Extraction providers `scripts/extract.py` auto-falls back across: 1. **Firecrawl** 2. **Linkup** 3. **Tavily** 4. **Exa** 5. **You.com** --- ## 🧠 Routing at a Glance Default priority (SerpBase excluded by design — opt-in only): ```text tavily → linkup → querit → exa →