Dev Browser
Give your coding agent a installed CLI to open pages, click, fill forms, screenshot, and scrape with session state across steps.
Overview
Dev Browser is an agent skill most often used in Build (also Ship and Validate) that exposes a CLI for persistent browser automation—navigation, forms, screenshots, scraping, and site testing.
Install
npx skills add https://github.com/sawyerhood/dev-browser --skill dev-browserWhat is this skill?
- Global npm CLI (dev-browser) with install bootstrap for local browser control
- Persistent page state across scripted navigation and interactions
- Sandboxed JavaScript scripts for clicks, forms, screenshots, and scraping
- Trigger-aligned for go-to URL, log in, automate, and test website requests
- Reference via dev-browser --help for command surface
- Two-step setup: npm install -g dev-browser then dev-browser install
Adoption & trust: 2.1k installs on skills.sh; 6.2k GitHub stars; 1/3 security scanners passed (skills.sh audits).
What problem does it solve?
You want your agent to drive a real browser for QA or data collection but one-off scripts drop state and break multi-step workflows.
Who is it for?
Solo builders who delegate repetitive browser tasks to an agent while building or smoke-testing web products.
Skip if: Teams that already maintain a full Playwright/Cypress suite and only need CI runners without agent-driven browsing.
When should I use this skill?
Users ask to navigate websites, fill forms, take screenshots, extract web data, test web apps, automate browser workflows, or say go to [url], click on, scrape, log into.
What do I get? / Deliverables
The agent runs dev-browser commands and sandboxed JS so pages stay in context through login, forms, captures, and automated checks.
- Browser screenshots
- Extracted web data
- Completed automated browser workflows
Recommended Skills
Journey fit
Spans multiple journey phases - primary shelf plus alternate fits below.
Browser control is core agent capability during product construction—wiring real UI flows and data extraction into builds. Fits agent-tooling because dev-browser is procedural glue for agents, not a one-off app feature.
Where it fits
Wire an agent to log into staging and walk through checkout while you implement the API.
Smoke-test the production landing page and capture screenshots after a deploy.
Click through a clickable prototype hosted on the web to confirm flows before full build.
Pull structured copy or pricing blocks from a competitor page for a content refresh.
How it compares
Agent-oriented browser CLI skill, not a hosted scraping API or a visual no-code recorder.
Common Questions / FAQ
Who is dev-browser for?
Indie and solo developers using Claude Code, Cursor, or Codex who need dependable browser automation from natural-language requests.
When should I use dev-browser?
During Build for integration and agent-tooling flows, during Validate to exercise prototypes, and during Ship when you ask the agent to test a website, screenshot pages, scrape data, or log into staging.
Is dev-browser safe to install?
It installs a global npm CLI with browser access; review the Security Audits panel on this page and treat credentials and production data carefully in automated sessions.
SKILL.md
READMESKILL.md - Dev Browser
# Dev Browser A CLI for controlling browsers with sandboxed JavaScript scripts. ## Installation ```bash npm install -g dev-browser dev-browser install ``` ## Usage Run `dev-browser --help` to learn more.