What problem does it solve? It gives an AI agent reliable access to current web information—latest facts, official API docs, and exact error messages—by querying official search engines and returning verbatim-verified evidence with source URLs instead of unverified summaries. ## Core Features & Use Cases - Dual-engine evidence search: Runs queries in parallel against Anthropic and OpenAI official web search, merges sources, and marks each quote as verified (✓) or unverified (?) against raw material. - Fast free lookup: A --quick mode via TinyFish returns a result list in about 2 seconds for finding accurate URLs or error text. - URL-to-Markdown fetching: Converts known URLs into clean Markdown body text, with support for llms.txt indexes and .md-suffixed LLM-friendly doc pages. - Use Case: When debugging an unfamiliar error, run a quick search to find the exact error thread, then fetch the official documentation page as Markdown to confirm the fix against the primary source. ## Quick Start Ask the agent to use the web-search skill to search for the latest documentation on a specific API and fetch the official page content.