skills/wraps-team/skills/wraps-cli/Gen Agent Trust Hub

wraps-cli

Fail

Audited by Gen Agent Trust Hub on Feb 16, 2026

Risk Level: CRITICALEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONCOMMAND_EXECUTIONCREDENTIALS_UNSAFEDATA_EXFILTRATION
Full Analysis
  • EXTERNAL_DOWNLOADS (CRITICAL): The skill instructs the agent to run @wraps.dev/cli using npx or npm install. This executes code from an untrusted source that is not part of the defined trust scope. Given the high privileges required (AWS Administrator access), this constitutes a critical risk of supply chain attack or arbitrary code execution.
  • REMOTE_CODE_EXECUTION (CRITICAL): The use of npx to execute a package directly from the npm registry allows for immediate execution of remote, unvetted code within the user's environment.
  • CREDENTIALS_UNSAFE (HIGH): The skill explicitly targets sensitive AWS credential files (~/.aws/credentials) and environment variables (AWS_SECRET_ACCESS_KEY). If the untrusted CLI is malicious, it has immediate access to these secrets.
  • COMMAND_EXECUTION (HIGH): The CLI is designed to perform high-impact operations including iam:CreateRole and iam:AttachRolePolicy. This provides a direct path for privilege escalation within the user's AWS account.
  • DATA_EXFILTRATION (MEDIUM): The CLI includes a telemetry feature that sends usage data to an external domain (wraps.dev) which is not whitelisted. While described as anonymous, this mechanism could be used to exfiltrate sensitive metadata or configuration data from the environment.
Recommendations
  • AI detected serious security threats
Audit Metadata
Risk Level
CRITICAL
Analyzed
Feb 16, 2026, 12:36 PM