test-driven-development
Pass
Audited by Gen Agent Trust Hub on Aug 14, 2026
Risk Level: SAFE
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill describes workflows involving the Chrome DevTools MCP to inspect live browser content, creating an attack surface for indirect prompt injection from malicious websites.
- Ingestion points: SKILL.md (Browser Testing with DevTools section) specifies reading DOM, console logs, and network requests.
- Boundary markers: The skill includes a dedicated 'Security Boundaries' section that explicitly warns the agent that browser content is untrusted data and must not be interpreted as instructions.
- Capability inventory: The agent uses DevTools for inspection and may spawn subagents to write reproduction tests as documented in SKILL.md.
- Sanitization: The skill instructs the agent to never navigate to extracted URLs without confirmation and to never access credentials via JavaScript execution.
Audit Metadata