What problem does it solve? When an AI agent needs to search the web or fetch a page, it often defaults to a single tool regardless of the task, wasting credits on expensive APIs or failing on JavaScript-heavy and anti-bot pages. This Skill provides explicit decision tables so the right search or scraping tool is chosen for each situation. ## Core Features & Use Cases - Search Routing: Directs queries to parallel-search for general lookups, Exa for semantic and find-similar discovery, and Perplexity for synthesized cited answers or deep research reports. - Fetch/Scrape Routing: Uses Jina Reader as the fast, free default for clean markdown, escalating to Firecrawl only for JS-heavy pages, anti-bot protection, structured extraction, or multi-page crawls. - Platform Deferral: Hands off LinkedIn and Reddit requests to the dedicated linkedin-scrape and reddit-scrape skills, since Jina, Firecrawl, and WebFetch are blocked on those domains. - Use Case: You ask the agent to research a competitor and pull content from their blog. The Skill routes the discovery queries to parallel-search, fetches the articles via Jina Reader, and escalates to Firecrawl only if a page renders as empty JavaScript. ## Quick Start Ask the agent to search for recent news about a company and fetch the top article, and it will pick the appropriate search and fetch tools automatically.
