Skills
skills/tsaol/awesome-claude/browser-use-agentcore/Gen Agent Trust Hub

browser-use-agentcore

Warn

Audited by Gen Agent Trust Hub on Feb 17, 2026

Risk Level: MEDIUMEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION] (MEDIUM): The scripts/browser_test.py file uses subprocess.check_call to execute pip install commands at runtime. This allows the skill to execute system-level commands to modify the environment without explicit user confirmation.\n- [EXTERNAL_DOWNLOADS] (MEDIUM): The skill automatically downloads and installs Python packages (bedrock-agentcore, playwright, nest_asyncio, boto3) via the ensure_dependencies function. Automated installation of packages at runtime is a security risk as it can be used for dependency injection or to install malicious code.\n- [PROMPT_INJECTION] (LOW): The skill is vulnerable to Indirect Prompt Injection (Category 8).\n
  • Ingestion points: The test_url function in scripts/browser_test.py ingests untrusted data from external websites via page.goto(url).\n
  • Boundary markers: None. There are no delimiters or instructions to ignore embedded commands in the processed web content.\n
  • Capability inventory: The script can perform network requests via Playwright and write files to the local system (screenshots).\n
  • Sanitization: None. Page titles and console errors are captured and returned directly to the agent without sanitization or escaping.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Feb 17, 2026, 06:40 PM