frontend-agent
Pass
Audited by Gen Agent Trust Hub on Mar 15, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: A thorough review of all skill components confirms the absence of obfuscated code, hidden instructions, or malicious intent.
- [COMMAND_EXECUTION]: The skill includes instructions to run standard development commands, such as TypeScript type checking (tsc) and unit testing (vitest), which are necessary for its primary purpose of frontend development.
- [PROMPT_INJECTION]: No patterns indicating attempts to override system prompts or bypass safety filters were detected. The instructions are focused on guiding the agent's technical output.
- [DATA_EXFILTRATION]: No hardcoded credentials or instructions to access sensitive local files (like .env or .ssh) were found. API integrations demonstrated in examples use appropriate local or relative endpoints.
- [EXTERNAL_DOWNLOADS]: The skill references standard development libraries and shadcn/ui components retrieved from official registries. It explicitly instructs the agent to report dependency needs rather than installing them independently, maintaining control over the environment.
Audit Metadata