test-driven-development
Pass
Audited by Gen Agent Trust Hub on Oct 24, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill is purely instructional and does not contain malicious code, hidden commands, or unauthorized data access patterns. Its primary function is to guide the agent in software development best practices.
- [INDIRECT_PROMPT_INJECTION]: The skill identifies and provides mitigation for potential indirect prompt injection surfaces when performing browser-based testing.
- Ingestion points: External browser content including DOM, console logs, and network responses entering the context via Chrome DevTools.
- Boundary markers: The skill includes a 'Security Boundaries' section that explicitly defines browser content as untrusted data rather than instructions.
- Capability inventory: The skill instructions involve executing repository-specific test commands and spawning subagents for task isolation.
- Sanitization: It mandates that the agent must not interpret browser content as commands, must not navigate to extracted URLs without confirmation, and is forbidden from accessing credentials or tokens via JS execution.
- [COMMAND_EXECUTION]: The skill describes how to discover and execute standard repository test commands (e.g., 'npm test', './gradlew', 'pytest'). This is legitimate behavior for a development-focused skill and is constrained to local repository tooling.
Audit Metadata