Browser Use API

Cloud browser automation via Browser Use API. Use when you need AI-driven web browsing, scraping, form filling, or multi-step web tasks without local browser control. Triggers on "browser use", "cloud browser", "scrape website", "automate web task", or when local browser isn't available/suitable.

Security Scan

The skill's code and instructions match its stated purpose (cloud browser automation) and only require a single API key, but it will send whatever you ask it to the remote api.browser-use.com service and the package assumes common CLI tools that aren't declared — review trust/privacy before use.

Purpose & Capability

Name/description (cloud browser automation) align with the actual behavior: the SKILL.md and helper script submit tasks to https://api.browser-use.com, poll status, and return results. The only declared requirement is BROWSER_USE_API_KEY, which is appropriate for a hosted API service.

Instruction Scope

Instructions and the helper script send task text to a third-party API and may include screenshots/outputs returned by that service. This is expected for a cloud browser tool, but it means any page contents, credentials, or sensitive data you include in a task will be transmitted to and processed by the remote service. SKILL.md does not provide privacy/retention details or warnings about sending PII/secrets.

Install Mechanism

There is no install spec (instruction-only plus a helper script), so nothing is downloaded or installed by the skill itself. The helper script uses curl and python3; no packages are pulled from external/unknown URLs by the skill.

Credentials

The only required env var is BROWSER_USE_API_KEY, which is proportionate. Minor inconsistency: the registry metadata lists no primary credential even though an API key is required. Also the script relies on curl and python3 (not declared in required binaries). The API key will be sent in an HTTP header to api.browser-use.com as part of normal operation.

Persistence & Privilege

The skill does not request permanent/always-on inclusion and does not modify other skills or global agent settings. It is user-invocable and can be invoked autonomously (platform default), which is expected.

Assessment

This skill appears to do what it says: it sends tasks to a remote cloud browser service and returns results. Before installing or using it: - Only use with an API key you trust: BROWSER_USE_API_KEY is required and will be sent to api.browser-use.com in request headers. - Do not include secrets or sensitive PII in task descriptions (URLs, login credentials, API keys, proprietary content), because those will be transmitted to and processed by the third-party service and may be captured in screenshots/steps. - Confirm the provider/domain (api.browser-use.com) and review their privacy/security policies if possible; the package has no homepage or provenance metadata. - The helper script expects curl and python3 on PATH even though the skill metadata doesn't list required binaries — ensure your environment provides these tools. - For high-volume or sensitive scraping, prefer local/browser-controlled tooling you manage, or test with non-sensitive queries first. If you need further verification, provide the service's homepage/owner info or any documentation that establishes who operates api.browser-use.com and how they handle data retention and security.

Browser Use

Cloud-based AI browser automation. Send a task in plain English, get structured results.

Quick Start

# Submit task
curl -s -X POST https://api.browser-use.com/api/v2/tasks \
  -H "X-Browser-Use-API-Key: $BROWSER_USE_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"task": "Go to example.com and extract the main heading"}'

# Poll for result (replace TASK_ID)
curl -s "https://api.browser-use.com/api/v2/tasks/TASK_ID" \
  -H "X-Browser-Use-API-Key: $BROWSER_USE_API_KEY"

Helper Script

Use scripts/browser-use.sh for simpler execution:

# Run task and wait for result
./scripts/browser-use.sh "Go to hacker news and get the top 3 stories"

# Just submit (don't wait)
./scripts/browser-use.sh --no-wait "Search Google for AI news"

API Reference

Create Task

POST https://api.browser-use.com/api/v2/tasks

Body:

{
  "task": "Plain English description of what to do",
  "llm": "gemini-3-flash-preview"  // optional, default is fast model
}

Response:

{
  "id": "task-uuid",
  "sessionId": "session-uuid"
}

Get Task Status

GET https://api.browser-use.com/api/v2/tasks/{taskId}

Response fields:

• status: pending | started | finished | failed
• output: Result text when finished
• steps: Array of actions taken (with screenshots)
• cost: Cost in dollars (e.g., "0.02")
• isSuccess: Boolean result

Stop Task

POST https://api.browser-use.com/api/v2/tasks/{taskId}/stop

Pricing

~$0.01-0.05 per task depending on complexity. Check balance:

curl -s https://api.browser-use.com/api/v2/credits \
  -H "X-Browser-Use-API-Key: $BROWSER_USE_API_KEY"

When to Use

• Complex multi-step web workflows
• Sites that block simple scraping
• Form filling and submissions
• When you need screenshots of steps
• When local browser control isn't available

When NOT to Use

• Simple page fetches (use web_fetch instead)
• When you have local browser access (use browser tool)
• Rapid/high-volume scraping (use Code Use or local scraping)