deploy
Warn
Audited by Gen Agent Trust Hub on Mar 10, 2026
Risk Level: MEDIUMCOMMAND_EXECUTIONREMOTE_CODE_EXECUTION
Full Analysis
- [COMMAND_EXECUTION]: The skill instructs the agent to perform destructive file operations using
rm -rfon thesrc/app/apiandsrc/app/configdirectories. While intended to resolve static export conflicts, such commands pose a risk of data loss if triggered incorrectly. - [COMMAND_EXECUTION]: The documentation recommends running a local script located at
./bin/setup.shfor environment configuration. Since the script's contents are not included in the skill, its behavior and safety cannot be verified. - [REMOTE_CODE_EXECUTION]: The skill directs the agent to use
npx -yto fetch and execute remote packages, specifically@modelcontextprotocol/server-githubandvercel-mcp-server, which involves the dynamic execution of external code.
Audit Metadata