playwright
Pass
Audited by Gen Agent Trust Hub on Mar 18, 2026
Risk Level: SAFEPROMPT_INJECTIONEXTERNAL_DOWNLOADSCOMMAND_EXECUTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: Fetches official Playwright libraries and browser binaries from Microsoft's infrastructure.
- [COMMAND_EXECUTION]: Provides standard CLI commands for environment setup and execution of browser automation tasks.
- [PROMPT_INJECTION]: The skill provides browser automation capabilities which expose an indirect prompt injection surface.
- Ingestion points: External web content is ingested through functions like
page.goto(),page.textContent(), andpage.evaluate()inSKILL.md. - Boundary markers: No delimiters or protective markers are used in the examples to isolate web content from agent instructions.
- Capability inventory: The agent can perform interactive actions like
page.click()andpage.fill()as demonstrated inSKILL.md. - Sanitization: No sanitization or data validation methods for external content are shown in the provided examples.
- [SAFE]: No malicious obfuscation, hidden commands, or unauthorized data exfiltration were detected in the skill documentation.
Audit Metadata