tinyfish

Automate web tasks from natural language via TinyFish MCP endpoints.

Updated Mar 29, 2026
One-click install
npx skills add https://github.com/sigitpoerwo/repoworkspace_zahra --skill tinyfish-sigitpoerwo
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: tinyfish
Source: https://github.com/sigitpoerwo/repoworkspace_zahra/tree/main/skills/05-CUSTOM-SKILLS/tinyfish
Command: npx skills add https://github.com/sigitpoerwo/repoworkspace_zahra --skill tinyfish-sigitpoerwo

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

TinyFish translates natural language instructions into automated web actions, enabling data extraction, form filling, and workflow automation against dynamic websites.

Core Features & Use Cases

  • Natural language control of web automation to perform scraping, form submissions, and UI interactions without writing selectors.
  • Use cases include data extraction, lead generation, monitoring, testing, and content aggregation across JS-heavy sites.
  • Quick example: describe a task like "extract product data from a site and fill a form" and TinyFish handles it end-to-end.

Quick Start

Start the TinyFish MCP server locally and begin using the web_scrape and web_automation tools.

Frequently Asked Questions about tinyfish

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

FAQPage Schema
How do I automate web scraping and form filling without writing CSS selectors?

Web automation from natural language allows you to perform scraping and form filling by describing tasks in plain text. You provide structured goals and URLs, and the system translates them into deterministic web interactions without requiring manual CSS selectors.

Can I extract data from JavaScript-heavy dynamic websites using natural language?

Yes, data extraction from JavaScript-heavy dynamic websites is supported. The automation handles dynamic site interactions, enabling scraping, lead generation, and content aggregation by interpreting natural language instructions to produce deterministic outputs.

How does natural language web automation work for UI interactions and testing?

Natural language web automation translates plain text instructions into automated web actions. You define structured goals and URLs, and the system executes web interactions for UI testing, form submissions, and data extraction to produce deterministic outputs.

Do I need a specific server environment to run natural language web automation?

Yes, running natural language web automation requires starting the TinyFish MCP server locally. The web_scrape and web_automation endpoints process structured goals and URLs to execute deterministic outputs against dynamic websites.

What is the best way to generate leads and aggregate content across multiple dynamic sites?

Using natural language web automation is an effective way to generate leads and aggregate content. By providing structured goals and URLs, you automate scraping and form submissions across dynamic sites to gather data deterministically without manual scripting.

What are the limitations of using natural language for web scraping and automation?

A key limitation is that natural language web scraping requires structured goals and specific URLs to produce deterministic outputs. Without clearly defined instructions and the local MCP server running, the automation cannot reliably execute web interactions.