

Firecrawl
Open source web data API for agents: search, scrape, crawl and interact, with an official MCP server
Open-source browser agent that finishes a web task you describe in plain words

About
Scout is installed from npm as `@outcrawl/scout` and has no built-in model: you assign models per role (hands for each step, an optional brain for planning, eyes for screenshots, extract for answers), each on any service that speaks the chat completions API, including a local server. Pass a JSON Schema with a task and the answer comes back in that shape. If it cannot finish, it says what is missing and returns what it found. It runs against any Chrome or Chromium on your own machine, or on Outcrawl's cloud, which adds residential-style browsers, captcha solving and a replay of every run. Each run reports steps and, when known, model spend. Repo: https://github.com/outcrawlai/scout.
Features
Install
I compare five ways to give an AI agent a browser, from local Chrome over MCP to hosted cloud browsers, with real install commands and pricing.
The APIs and cloud browsers that let an agent search, read and use the live web, compared on what they return, how they bill and how to plug them in.
Six building blocks that turn a chat model into an agent that can work on its own: an inbox, memory, a safe place to run code, app access, background jobs and tracing.



Free. I review every submission.
is in the review queue. I'll install it, test it and list it if it passes.