Skip to content

AWS AgentCore Cloud Browser

Run agent-browser automation on AWS Bedrock AgentCore cloud browsers using your AWS credentials

Stars
45,202
Licence
Apache-2.0
Files
1 file

What it does

This skill teaches an AI agent to run its browser automation on a managed cloud browser hosted by AWS Bedrock AgentCore instead of a local Chrome. The agent uses the same agent-browser commands (open, snapshot, click, screenshot) and only changes where the browser runs. It picks up AWS credentials from environment variables or the AWS CLI, and can keep login state between sessions using a persistent profile.

Great for

  • Developers whose browser automation must run on AWS infrastructure
  • Teams already using AWS credentials, SSO or IAM roles who want managed cloud browsers
  • Tasks that need a long-running or persistent logged-in browser session

Skip it if

  • Users without an AWS account or working AWS credentials, since the skill cannot authenticate without them
  • Simple local browsing where a cloud browser adds cost and an AWS dependency for no benefit
  • Handling highly sensitive logins through a persistent profile on shared infrastructure, because stored cookies and sessions persist in AWS

Try asking

  • “Use AgentCore to open example.com and take a screenshot of the homepage”
  • “Run this login flow on AWS cloud browser and save the session in profile my-app”
  • “Open the dashboard in eu-west-1 using our AWS SSO profile and summarise the figures”
Step-by-step guide: how to use AWS AgentCore Cloud BrowserInstall, first run, examples, tips and FAQs

Install in 3 steps

  1. 1

    Download the skill

    The ZIP contains the 1 original files, the Apache-2.0 licence and an attribution note.

    agentcore.zip
  2. 2

    Add it to your agent

    Claude Code (all projects)

    mkdir -p ~/.claude/skills && unzip agentcore.zip -d ~/.claude/skills/

    Claude Code / Agent SDK (this project only)

    mkdir -p .claude/skills && unzip agentcore.zip -d .claude/skills/

    Claude apps (web/desktop): Settings → Capabilities → Skills → Upload skill, and choose the ZIP.

  3. 3

    Ask for the task

    No special command needed — the agent reads the skill's description and loads it when your request matches. Try one of the prompts above.

Get better results

  • Use a dedicated least-privilege IAM role or SSO profile rather than long-lived root keys, and avoid hardcoding keys in scripts.
  • Set AGENTCORE_SESSION_TIMEOUT for longer tasks, and close sessions with agent-browser close when finished.
  • Open the Live View URL printed at session start to watch or check what the agent is doing in real time.

What's inside

  • SKILL.md

Redistributed unmodified from vercel-labs/agent-browser under the Apache-2.0 licence. The assessment was drafted with AI from the skill's files on 2026-10-29. Skills can instruct your agent to run code — read SKILL.md and any scripts before installing.