alibabacloud-emr-starrocks-assistant

Pass

Audited by Gen Agent Trust Hub on May 29, 2026

Risk Level: SAFE
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill's diagnostic tool sr-doctor fetches the client's public egress IP from ipinfo.io (a well-known technology service) to provide the user with the correct CIDR for cluster whitelisting. This operation is performed neutrally and is restricted to IP discovery.
  • [COMMAND_EXECUTION]: The skill uses a bundled Python CLI suite (srsql, sr-login, sr-whoami, sr-doctor) to interact with StarRocks clusters. These commands are local to the skill and are used to provide a structured interface for SQL execution and cluster diagnostics.
  • [CREDENTIALS_UNSAFE]: Cluster credentials (user and password) are stored locally in ~/.starrocks/{profile}.cnf. The skill implements industry-standard security practices for local storage by enforcing 0700 directory permissions and 0600 file permissions to prevent unauthorized access by other users on the system.
  • [DATA_EXFILTRATION]: The skill documentation includes a section on sensitive data masking, ensuring that API keys (AK), passwords in error messages, and query results containing tokens are truncated or masked before being displayed to the user.
  • [PROMPT_INJECTION]: The SKILL.md file contains clear instructions for the AI agent to follow a specific bootstrap protocol and to gate non-READ SQL statements behind a user-confirmation flag (--yes). These instructions are designed to enhance operational safety rather than bypass it.
Audit Metadata
Risk Level
SAFE
Analyzed
May 29, 2026, 03:05 AM
Security Audit — agent-trust-hub — alibabacloud-emr-starrocks-assistant