surf-search-skill

Route web search and extraction across Tavily, Parallel AI, and Brave with key rotation.

3|1|Updated May 1, 2026
One-click install
npx skills add https://github.com/frederico-kluser/surf-skill --skill surf-search-skill
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: surf-search-skill
Source: https://github.com/frederico-kluser/surf-skill/tree/main
Command: npx skills add https://github.com/frederico-kluser/surf-skill --skill surf-search-skill

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

Web research breaks agent workflows when a provider is down, a key is invalid, or the agent times out—surf-search-skill solves this by routing requests across multiple web providers with automatic key rotation, retries, and last-known-good persistence.

Core Features & Use Cases

  • Multi-provider web access: Executes search, extract, crawl, URL mapping, and deep research through Tavily and Parallel AI (with Brave search as a fallback).
  • Automatic key rotation + fallback: Detects auth/billing failures, burns failing keys, retries on rate limits, and falls back to the next provider/key while remembering the last successful provider.
  • Agent-friendly outputs: Produces clean Markdown by default, supports a normalized JSON envelope, and emits stable progress logs to stderr for orchestration.

Quick Start

Use the skill to search the web and gather sources by asking: surf-search-skill search "tavily vs parallel ai for agent research" --max 3

Frequently Asked Questions about surf-search-skill

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

FAQPage Schema
How do I handle web search API failures in AI agent workflows?

Web search API failures in agent workflows can be handled by routing requests across multiple providers with automatic key rotation, retries, and provider fallback to maintain continuous research operations.

How does API key rotation work for web scraping and search?

API key rotation for web search discovers available keys, burns failing keys on auth or billing failures, retries on rate limits, and falls back to the next available provider while remembering the last successful one.

Can I use Tavily and Brave search together for web content extraction?

Yes, you can use Tavily and Brave search together by routing web search and content extraction requests across both providers with Parallel AI, using Brave as a fallback when primary providers fail.

What's the best way to orchestrate async research workflows for coding agents?

Async research workflows for coding agents are best orchestrated using batched queries with stable progress logging to stderr, normalized JSON envelopes, and clean Markdown outputs for reliable agent processing.

Why does my agent web research timeout when a provider goes down?

Agent web research times out when a provider goes down because single-provider setups lack fallback mechanisms; routing across multiple web providers with last-known-good persistence prevents these workflow interruptions.

Does surf-search-skill support URL mapping and crawling for research planning?

Yes, surf-search-skill supports URL mapping and crawling alongside search and extraction operations, enabling comprehensive research planning with optional TTL-based caching to guard expensive operations.