Unsandboxed browser automation
HighAV-RUNTIME-008Runtime safetyCore pack
What it finds
Puppeteer/Playwright/Selenium usage combined with `page.evaluate(`, `--no-sandbox`, or `dangerouslySetInnerHTML`.
Why it matters
An agent driving an unsandboxed browser can be steered by content on visited pages — a live path from prompt injection to real clicks and downloads.
How to fix it
Run browser automation sandboxed with no persisted credentials; treat all page content as untrusted.
Pro shows the corrected code for your exact file, step by step, and re-scans to confirm the fix. See plans
Check details
- Finding code
- UNSAFE_BROWSER_AUTOMATION
- How it detects
- Looks for a risky pattern, and stays quiet when a matching safeguard is present.
- Confidence
- Likely: strong signal, but review it in context.
- Scans
- Puppeteer/Playwright/Selenium automation source
Check your agent for this
Paste your agent code or MCP config. The free scan runs all 124 checks in about 10 seconds — no account needed.
Related checks
- Command injection riskAV-RUNTIME-004 · Critical
- Unsafe dynamic code executionAV-RUNTIME-005 · Critical
- Privileged or host-mounted container configurationAV-RUNTIME-007 · Critical
- SQL query built from unescaped inputAV-RUNTIME-009 · Critical
- Unsafe deserialization of untrusted dataAV-RUNTIME-010 · Critical
- Prompt injection surface detectedAV-RUNTIME-002 · High