check-agentkit-prod

Warn

Audited by Gen Agent Trust Hub on Aug 28, 2026

Risk Level: MEDIUMCREDENTIALS_UNSAFECOMMAND_EXECUTIONDATA_EXFILTRATION
Full Analysis
  • [CREDENTIALS_UNSAFE]: The skill instructs the agent to execute echo $SCALEKIT_CLIENT_SECRET in Step 2, which causes sensitive production credentials to be displayed in the clear within the agent's output history and session logs.
  • [DATA_EXFILTRATION]: In Step 2, the skill uses curl to send SCALEKIT_CLIENT_ID and SCALEKIT_CLIENT_SECRET as POST data to the address stored in SCALEKIT_ENVIRONMENT_URL. While intended for the vendor's API, this pattern transmits production secrets over the network and relies on a user-provided environment variable to define the destination.
  • [COMMAND_EXECUTION]: The skill utilizes shell commands (echo, curl, rg) to verify environment configurations, test API connectivity, and scan the local filesystem for hardcoded secrets.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Aug 28, 2026, 08:09 AM
Security Audit — agent-trust-hub — check-agentkit-prod