huawei-cloud-rds-smart-service
Pass
Audited by Gen Agent Trust Hub on Aug 19, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The
references/cli-installation-guide.mdfile provides commands to download and execute the officialhcloudCLI installation script from a Huawei Cloud domain (obs.cn-north-4.myhuaweicloud.com). This is a standard and expected deployment method for vendor-provided cloud tooling. - [COMMAND_EXECUTION]: The skill uses shell commands and Python scripts to interface with the Huawei Cloud API. The
tests/run_tests.pyscript utilizessubprocess.runwithshell=Trueto execute test commands defined intest_cases.json. While this is common in CLI testing utilities, it represents a potential command injection surface if the parameters, such as instance IDs, are sourced from untrusted inputs without sanitization. - [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and analyze external database data, including slow query logs and error logs. These logs could potentially contain malicious prompts injected by database users intended to influence the agent's behavior during analysis.
- Ingestion points: Data enters via log listing operations (
ListSlowLogs,ListErrorLogs,ListAuditlogs) as described inSKILL.mdandreferences/api-paths.md. - Boundary markers: The skill does not implement specific delimiters for processed log data but requires explicit human confirmation for any action that modifies cloud resources.
- Capability inventory: The skill possesses capabilities for instance management (restart, resize, disk volume changes), parameter tuning, and backup recovery across its script library.
- Sanitization: There is no evidence of log content filtering or prompt injection sanitization in the provided script handlers.
- [SAFE]: The skill adheres to security best practices by recommending the use of environment variables for authentication (
HUAWEI_ACCESS_KEY,HWC_AK, etc.) rather than hardcoding secrets. Furthermore, the inclusion of a detailedreferences/iam-policies.mdguide promotes the principle of least privilege for the agent's execution environment.
Audit Metadata