retellai-incident-runbook
Pass
Audited by Gen Agent Trust Hub on Sep 9, 2026
Risk Level: SAFEDATA_EXFILTRATIONCOMMAND_EXECUTION
Full Analysis
- [DATA_EXFILTRATION]: The skill provides instructions to extract and decode sensitive API credentials from the infrastructure environment.
- Evidence: In
references/response-procedures.md, the commandkubectl get secret retellai-secrets -o jsonpath='{.data.api-key}' | base64 -dis intended to verify the API key. This action decodes the secret into the agent's output buffer, potentially exposing the raw credential to the user or recording it in session logs. - [COMMAND_EXECUTION]: The skill provides procedures for the agent to perform administrative operations on production infrastructure.
- Evidence: The runbook includes instructions for modifying deployment environment variables (
kubectl set env) and managing secrets (kubectl create secret), which allows the agent to alter the operational state of the telephony integration.
Audit Metadata