auth-tool-cloudbase
Pass
Audited by Gen Agent Trust Hub on Oct 23, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill interacts with external cloud environment configurations, creating a surface where malicious instructions could theoretically be embedded in environment metadata or provider settings.
- Ingestion points: Environment configuration and provider status are fetched via
queryAppAuthandcallCloudApi(references/extended-guide.md). - Boundary markers: The skill includes explicit warnings about the security trade-offs of anonymous login and references a 'Change Safety Protocol' to guide modifications (SKILL.md, references/extended-guide.md).
- Capability inventory: The skill can modify authentication strategies, update third-party provider credentials (SMTP, OAuth), and manage environment API keys via
manageAppAuthandcallCloudApi(references/extended-guide.md). - Sanitization: No specific input sanitization is defined for the parameters passed to the MCP management tools.
- [COMMAND_EXECUTION]: The skill is designed to execute administrative commands and API calls to manage the lifecycle of authentication resources.
- Evidence: Utilizes
manageAppAuthandcallCloudApito perform actions likeModifyProvider,ModifyLoginConfig, andCreateApiKey(references/extended-guide.md).
Audit Metadata