chrome-automation
The chrome-automation skill provides a command-line interface for agents to interact with a user's active Chrome browser session. It supports navigation, form filling, data extraction, and replaying DevTools recordings by connecting to a running Chrome instance with remote debugging enabled.
Is chrome-automation safe to install?
Review the source first: our audit of chrome-automation's source files found 26 shell commands, 0 external URLs, file reads and writes (high risk). Every command and URL listed appears verbatim in the skill's source. The skill executes shell commands to control the user's browser, which can perform destructive actions like deleting data or publishing content. It also reads local files for recording replays and platform-specific references.
How we audit skills: our security review methodology.
Who is this skill for?
Developers and AI agents requiring browser automation capabilities within a user's existing, authenticated Chrome session.
What can you do with it?
- Automating social media posting and platform management
- Extracting structured data from web pages
- Replaying recorded browser workflows
- Filling forms and interacting with web applications
- Performing multi-platform tasks using sequential subagents
How good is this skill?
Quality score: 5/10. The documentation is comprehensive, providing clear command references, troubleshooting steps for common issues like iframes, and a structured approach for multi-platform automation.
What does the skill file contain?
# Skill: Chrome Automation (agent-browser) Automate browser tasks in the user's real Chrome session via the [agent-browser](https://github.com/vercel-labs/agent-browser) CLI. > **Prerequisite**: agent-browser must be installed and Chrome must have remote debugging enabled. See `references/agent-browser-setup.md` if unsure. --- ## Core Principle: Reuse the User's Existing Chrome This skill operates on a **single Chrome process** — the user's real browser. There is no session management, no separate profiles, no launching a fresh Playwright browser. ### Always Start by Listing Tabs Before...
Frequently asked questions
Does this skill launch a new browser instance?
No. It connects to the user's existing Chrome process using the --auto-connect flag to preserve cookies and login sessions.
How does the skill handle iframes?
The snapshot command cannot see inside iframes. The skill recommends using the eval command to access iframe content or keyboard commands for blind input.
What should I do if a task requires multiple platforms?
Launch sequential subagents, one for each platform, to manage context limits and avoid tab conflicts.
How do I verify that an action was successful?
Use snapshot -i to check the interactive state or screenshot to visually confirm the page status after each action.
Related skills
firecrawl-lead-gen
12.8KSales professionals, recruiters, and researchers who need to compile prospect lists from public web sources
The firecrawl-lead-gen skill extracts prospect information from web directories and databases using the Firecrawl browser. It generates structured lead lists in JSON, CSV, or Markdown formats based on user-defined criteria.
firecrawl-seo-audit
12.8KSEO specialists, content marketers, and website owners who need automated, data-driven site audits
The firecrawl-seo-audit skill performs website SEO audits by mapping site structures, scraping key pages, and comparing content against target keywords and competitor search results.
fetching-salesforce-docs
1.5KAI agents requiring authoritative, grounded information from Salesforce-owned documentation domains
A retrieval skill for fetching and grounding AI agent responses in official Salesforce documentation. It provides a playbook for navigating developer, help, architect, and design system portals while handling JS-heavy or shell-rendered pages.
seo-audit
Users requiring technical SEO audits, site health checks, or diagnostics for ranking and traffic issues
The SEO Audit skill performs technical and on-page SEO diagnostics using the Bright Data CLI to fetch and analyze live, JavaScript-rendered web content. It identifies issues related to schema markup, hreflang, canonicals, and search engine indexing, providing actionable reports with evidence-based findings.