huawei-cloud-lts-manage
Pass
Audited by Gen Agent Trust Hub on Sep 7, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill provides instructions for downloading the
hcloudCLI installation script from an official Huawei Cloud domain (obs.cn-north-4.myhuaweicloud.com). This is documented as the standard distribution channel for the vendor's command-line interface. - [REMOTE_CODE_EXECUTION]: The installation guide includes a pattern for downloading and immediately executing a shell script (
bash hcloud_install.sh). While this is a standard installation procedure for this vendor, it represents a remote code execution vector. - [COMMAND_EXECUTION]: The skill operates by executing shell commands via the
hcloudCLI. It correctly implements operational safety measures by requiring the agent to obtain explicit user confirmation before executing any delete operations. - [INDIRECT_PROMPT_INJECTION]: The skill interpolates user-supplied strings, such as search keywords, log group names, and SQL expressions, directly into shell command arguments. This creates a surface where untrusted data could potentially manipulate the intended CLI command execution.
- Ingestion points: Multiple parameters in
SKILL.md(e.g.,{keywords},{sql_expression},{log_group_name}) and test case definitions intemplates/test-vars.json. - Boundary markers: None. The skill does not use specific delimiters or instructions to isolate user-provided inputs within the command strings.
- Capability inventory: The skill has broad capabilities to list, create, update, and delete log groups, streams, transfers, and alarm rules.
- Sanitization: There is no evidence of parameter validation, escaping, or sanitization within the skill's instructions or scripts.
Audit Metadata