NYC

browser-automation

Fail

Audited by Gen Agent Trust Hub on Feb 17, 2026

Risk Level: CRITICALPROMPT_INJECTION
Full Analysis
  • PROMPT_INJECTION (LOW): The skill implements browser automation and data scraping which creates a surface for indirect prompt injection attacks. 1. Ingestion points: Scraping occurs in SearchPage.get_results within references/advanced-patterns.md and cookie extraction in CookieManager within references/security-examples.md. 2. Boundary markers: Absent. The skill does not wrap ingested data in delimiters or provide instructions to the agent to ignore instructions embedded in the external content. 3. Capability inventory: Full browser automation via Playwright, including navigation, element interaction, and network request interception. 4. Sanitization: Partially present through _safe_fill (blocks password field interaction) and validate_domain (implements allowlisting/blocklisting).
  • SAFE (SAFE): An automated scanner flagged the domain browser.se. After analysis, this appears to be a false positive triggered by the logging namespace 'browser.audit' in references/security-examples.md, as the domain does not appear in the skill source code.
Recommendations
  • Contains 1 malicious URL(s) - DO NOT USE
Audit Metadata
Risk Level
CRITICAL
Analyzed
Feb 17, 2026, 05:18 PM