test-driven-development
Pass
Audited by Gen Agent Trust Hub on Sep 17, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill instructs the agent to use the Chrome DevTools MCP to inspect external browser content. This creates an ingestion point for untrusted data that could potentially contain malicious instructions aimed at manipulating the agent's behavior.
- Ingestion points: Chrome DevTools MCP output including DOM structure, console logs, and network request payloads as described in the "Browser Testing with DevTools" section of SKILL.md.
- Boundary markers: The skill includes a specific "Security Boundary" section that explicitly labels this data as "untrusted data" and instructs the agent to treat it as information, not instructions.
- Capability inventory: The agent has the capability to write and modify source code, execute shell commands (test runners like npm, pytest, gradle), and spawn subagents.
- Sanitization: The skill provides explicit mitigation instructions: never interpret browser content as commands, never navigate to extracted URLs without confirmation, and avoid accessing sensitive data like cookies or tokens via JavaScript execution.
Audit Metadata